HOMO-LUMO Energy Gap Prediction using Molecular descriptors.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "id": "kIubEW7jdlu3",
      "metadata": {
        "id": "kIubEW7jdlu3"
      },
      "source": [
        "# HOMO-LUMO Energy Gap Prediction using RDKit Molecular descriptors<br>\n",
        "<b>HOMO</b> stands for <i>highest occupied molecular orbital</i> and <b>LUMO</b> stands for <i> lowest unoccupied molecular orbital</i>. HOMO-LUMO molecular orbitals are called frontier molecular orbitals. They are involved in chemical bond formation. Especially, pericyclic reactions such as cycloaddition, electrocyclic reactions, and sigmatropic rearrangement are explained using HOMO-LUMO molecular orbitals. In addition, in UV-Visible spectroscopy, the absorbance of organic molecules that have extended conjugated double bonds can be rationalized using the HOMO-LUMO energy gap of the molecules.\n",
        "<div>\n",
        "    <img src=\"https://github.com/gashawmg/RDKit-descriptors-for-HOMO-LUMO-energy-gap-prediction/blob/main/plainHOMO-LUMOEnegygap%20.jpg?raw=1\" width=\"400\")>\n",
        "</div>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 62,
      "id": "KMCvaSpIdlu6",
      "metadata": {
        "id": "KMCvaSpIdlu6"
      },
      "outputs": [],
      "source": [
        "from rdkit.Chem import AllChem\n",
        "from rdkit import Chem\n",
        "from rdkit.Chem import Descriptors\n",
        "from rdkit.ML.Descriptors import MoleculeDescriptors\n",
        "from rdkit.Chem import Draw\n",
        "from rdkit.Chem.Draw import IPythonConsole\n",
        "#----------------------------------------------------\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "#----------------------------------------------------\n",
        "from sklearn.model_selection import cross_val_score,train_test_split\n",
        "from sklearn.metrics import r2_score,mean_absolute_error, mean_squared_error\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "#----------------------------------------------------\n",
        "from matplotlib import pyplot as plt\n",
        "import matplotlib.patches as mpatches\n",
        "import seaborn as sn\n",
        "#----------------------------------------------------\n",
        "# https://github.com/shankarpandala/lazypredict\n",
        "import lazypredict\n",
        "from lazypredict.Supervised import LazyRegressor"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 63,
      "id": "GFJtR_Epdlu7",
      "metadata": {
        "id": "GFJtR_Epdlu7"
      },
      "outputs": [],
      "source": [
        "# import session_info\n",
        "# session_info.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "h1RtzNJLdlu8",
      "metadata": {
        "id": "h1RtzNJLdlu8"
      },
      "source": [
        "# Part I - Generate Molecular Descriptors from SMILES"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 64,
      "id": "EZxbEgMydlu8",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EZxbEgMydlu8",
        "outputId": "f66260b9-5484-48f1-b23f-de450c8eb9e8"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "(2904, 2)"
            ]
          },
          "execution_count": 64,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# https://zivgitlab.uni-muenster.de/m_kueh11/fp-dm-tool\n",
        "# File with SMILES and HOMO-LUMO energy gap values\n",
        "# Originally an excel file, but converted it to csv file \n",
        "# https://www.sciencedirect.com/science/article/pii/S2451929420300851\n",
        "\n",
        "dataset = pd.read_csv('Orbital_Energies_input_data.csv')\n",
        "dataset.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 65,
      "id": "jZ-JoEE0dlu9",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "jZ-JoEE0dlu9",
        "outputId": "d01bc0da-8964-4d0a-982b-75e69387fd62"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>SMILES</th>\n",
              "      <th>Energygap</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Cc1ccc(cc1)C(F)(F)F</td>\n",
              "      <td>197.75</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>OC(=O)CCCCl</td>\n",
              "      <td>247.49</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>CC(C)(Oc1ccc(CCNC(=O)c2ccc(Cl)cc2)cc1)C(=O)O</td>\n",
              "      <td>164.71</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Nc1ccc(Cl)c(Cl)c1</td>\n",
              "      <td>169.03</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>C[C@@H](CCO)CCC=C(C)C</td>\n",
              "      <td>209.57</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                         SMILES  Energygap\n",
              "0                           Cc1ccc(cc1)C(F)(F)F     197.75\n",
              "1                                   OC(=O)CCCCl     247.49\n",
              "2  CC(C)(Oc1ccc(CCNC(=O)c2ccc(Cl)cc2)cc1)C(=O)O     164.71\n",
              "3                             Nc1ccc(Cl)c(Cl)c1     169.03\n",
              "4                         C[C@@H](CCO)CCC=C(C)C     209.57"
            ]
          },
          "execution_count": 65,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "dataset.head()"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "drhHOj5Ydlu9",
      "metadata": {
        "id": "drhHOj5Ydlu9"
      },
      "source": [
        "## 1. Generate canonical SMILES"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 66,
      "id": "v24h0MMFdlu9",
      "metadata": {
        "id": "v24h0MMFdlu9"
      },
      "outputs": [],
      "source": [
        "# There might be one or more valid SMILES that can represent one compound\n",
        "def canonical_smiles(smiles):\n",
        "    mols = [Chem.MolFromSmiles(smi) for smi in smiles] \n",
        "    smiles = [Chem.MolToSmiles(mol) for mol in mols]\n",
        "    return smiles"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 67,
      "id": "xC4qSFnjdlu-",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 167
        },
        "id": "xC4qSFnjdlu-",
        "outputId": "0458019f-9833-4e12-f1ad-beb1392838c8"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<rdkit.Chem.rdchem.Mol at 0x233cb199850>"
            ]
          },
          "execution_count": 67,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "a = Chem.MolFromSmiles('C=CCC')\n",
        "a"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 68,
      "id": "5E7X9itLdlu-",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 167
        },
        "id": "5E7X9itLdlu-",
        "outputId": "63fb006a-542a-43dc-dc77-ca04c8491eb7"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<rdkit.Chem.rdchem.Mol at 0x233cb1999a0>"
            ]
          },
          "execution_count": 68,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "b = Chem.MolFromSmiles('CCC=C')\n",
        "b"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 69,
      "id": "QRixBcumdlu_",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QRixBcumdlu_",
        "outputId": "f6f10f89-6eee-41e2-dcd2-89786675e110"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "False"
            ]
          },
          "execution_count": 69,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "a==b"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 70,
      "id": "OnsXU-pMdlu_",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "OnsXU-pMdlu_",
        "outputId": "d544d819-ea05-4c7d-b52d-a526415bfe18"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "['C=CCC']"
            ]
          },
          "execution_count": 70,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "c = canonical_smiles(['C=CCC'])\n",
        "c"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 71,
      "id": "k4R0HqRfdlu_",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "k4R0HqRfdlu_",
        "outputId": "b6159c14-c3bf-43d8-ee72-228da01e7632"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "['C=CCC']"
            ]
          },
          "execution_count": 71,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "d = canonical_smiles(['CCC=C'])\n",
        "d"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 72,
      "id": "-HuCexrWdlu_",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-HuCexrWdlu_",
        "outputId": "8522d321-fb21-4646-9370-b3b7b69737e7"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "True"
            ]
          },
          "execution_count": 72,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "c==d"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 73,
      "id": "rW-YACdSdlvA",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rW-YACdSdlvA",
        "outputId": "67d4b9bb-a452-4a44-f331-59acca2b638a"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "2904"
            ]
          },
          "execution_count": 73,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# Canonical SMILES\n",
        "Canon_SMILES = canonical_smiles(dataset.SMILES)\n",
        "len(Canon_SMILES)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 74,
      "id": "eMun7uiedlvA",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "id": "eMun7uiedlvA",
        "outputId": "aaecfc5f-244d-4508-cd92-95a3182c4884"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>SMILES</th>\n",
              "      <th>Energygap</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Cc1ccc(C(F)(F)F)cc1</td>\n",
              "      <td>197.75</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>O=C(O)CCCCl</td>\n",
              "      <td>247.49</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>CC(C)(Oc1ccc(CCNC(=O)c2ccc(Cl)cc2)cc1)C(=O)O</td>\n",
              "      <td>164.71</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Nc1ccc(Cl)c(Cl)c1</td>\n",
              "      <td>169.03</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>CC(C)=CCC[C@@H](C)CCO</td>\n",
              "      <td>209.57</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2899</th>\n",
              "      <td>c1ccc(P(CCP(c2ccccc2)c2ccccc2)c2ccccc2)cc1</td>\n",
              "      <td>168.65</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2900</th>\n",
              "      <td>Brc1cccc2sccc12</td>\n",
              "      <td>162.93</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2901</th>\n",
              "      <td>CCOC(=O)N1c2ccccc2C=C[C@@H]1OCC</td>\n",
              "      <td>165.10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2902</th>\n",
              "      <td>c1ccc2sccc2c1</td>\n",
              "      <td>167.96</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2903</th>\n",
              "      <td>Cc1cccc(C)c1O</td>\n",
              "      <td>188.37</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2904 rows × 2 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "                                            SMILES  Energygap\n",
              "0                              Cc1ccc(C(F)(F)F)cc1     197.75\n",
              "1                                      O=C(O)CCCCl     247.49\n",
              "2     CC(C)(Oc1ccc(CCNC(=O)c2ccc(Cl)cc2)cc1)C(=O)O     164.71\n",
              "3                                Nc1ccc(Cl)c(Cl)c1     169.03\n",
              "4                            CC(C)=CCC[C@@H](C)CCO     209.57\n",
              "...                                            ...        ...\n",
              "2899    c1ccc(P(CCP(c2ccccc2)c2ccccc2)c2ccccc2)cc1     168.65\n",
              "2900                               Brc1cccc2sccc12     162.93\n",
              "2901               CCOC(=O)N1c2ccccc2C=C[C@@H]1OCC     165.10\n",
              "2902                                 c1ccc2sccc2c1     167.96\n",
              "2903                                 Cc1cccc(C)c1O     188.37\n",
              "\n",
              "[2904 rows x 2 columns]"
            ]
          },
          "execution_count": 74,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# Put the smiles in the dataframe\n",
        "dataset['SMILES'] = Canon_SMILES\n",
        "dataset"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 75,
      "id": "bHwem8_VdlvA",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bHwem8_VdlvA",
        "outputId": "899d61bd-a864-4d8a-da32-4069bb5d841f"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "31"
            ]
          },
          "execution_count": 75,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# Create a list for duplicate smiles\n",
        "duplicates_smiles = dataset[dataset['SMILES'].duplicated()]['SMILES'].values\n",
        "len(duplicates_smiles)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 76,
      "id": "OdLAXT4sdlvA",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "OdLAXT4sdlvA",
        "outputId": "b45ce3f7-596a-4881-e3ba-d9055c43a588"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>SMILES</th>\n",
              "      <th>Energygap</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>218</th>\n",
              "      <td>C1=CCCCCCC1</td>\n",
              "      <td>214.10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2125</th>\n",
              "      <td>C1=CCCCCCC1</td>\n",
              "      <td>214.10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1901</th>\n",
              "      <td>C1CC[C@H]2CCCC[C@@H]2C1</td>\n",
              "      <td>251.14</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>552</th>\n",
              "      <td>C1CC[C@H]2CCCC[C@@H]2C1</td>\n",
              "      <td>251.14</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>554</th>\n",
              "      <td>C=CN1CCCC1=O</td>\n",
              "      <td>190.83</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1808</th>\n",
              "      <td>C=CN1CCCC1=O</td>\n",
              "      <td>190.84</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2203</th>\n",
              "      <td>C=Cc1ccccc1</td>\n",
              "      <td>168.47</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1354</th>\n",
              "      <td>C=Cc1ccccc1</td>\n",
              "      <td>168.48</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2341</th>\n",
              "      <td>C=Cc1ccccc1</td>\n",
              "      <td>168.49</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1265</th>\n",
              "      <td>C=Cc1ccncc1</td>\n",
              "      <td>174.83</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>265</th>\n",
              "      <td>C=Cc1ccncc1</td>\n",
              "      <td>174.85</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2400</th>\n",
              "      <td>CC(C)(C)OC(=O)N1CCC(C(=O)O)CC1</td>\n",
              "      <td>208.68</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1621</th>\n",
              "      <td>CC(C)(C)OC(=O)N1CCC(C(=O)O)CC1</td>\n",
              "      <td>207.40</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1184</th>\n",
              "      <td>CC(C)C[C@@H](N)C(=O)O</td>\n",
              "      <td>224.37</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1759</th>\n",
              "      <td>CC(C)C[C@@H](N)C(=O)O</td>\n",
              "      <td>224.48</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2324</th>\n",
              "      <td>CC(C)I</td>\n",
              "      <td>192.15</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>85</th>\n",
              "      <td>CC(C)I</td>\n",
              "      <td>192.15</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2885</th>\n",
              "      <td>CC(C)[C@@H](N)C(=O)O</td>\n",
              "      <td>225.26</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2408</th>\n",
              "      <td>CC(C)[C@@H](N)C(=O)O</td>\n",
              "      <td>225.27</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>21</th>\n",
              "      <td>CC(C)[C@@H](N)C(=O)O</td>\n",
              "      <td>225.28</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1856</th>\n",
              "      <td>CC(C)[C@@H](N)CO</td>\n",
              "      <td>233.36</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2813</th>\n",
              "      <td>CC(C)[C@@H](N)CO</td>\n",
              "      <td>233.36</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1841</th>\n",
              "      <td>CC(C)[C@@H]1CC[C@@H](C)C[C@H]1O</td>\n",
              "      <td>239.02</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1448</th>\n",
              "      <td>CC(C)[C@@H]1CC[C@@H](C)C[C@H]1O</td>\n",
              "      <td>239.01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1881</th>\n",
              "      <td>CC1(C)[C@H]2CC=C(CO)[C@@H]1C2</td>\n",
              "      <td>204.96</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>819</th>\n",
              "      <td>CC1(C)[C@H]2CC=C(CO)[C@@H]1C2</td>\n",
              "      <td>204.96</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1635</th>\n",
              "      <td>CCCCCCCC/C=C/CCCCCCCC(=O)O</td>\n",
              "      <td>214.11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2310</th>\n",
              "      <td>CCCCCCCC/C=C/CCCCCCCC(=O)O</td>\n",
              "      <td>216.25</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>262</th>\n",
              "      <td>CI</td>\n",
              "      <td>195.30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>853</th>\n",
              "      <td>CI</td>\n",
              "      <td>195.29</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2061</th>\n",
              "      <td>CO[C@H]1O[C@H](CO)[C@H](O)[C@H](O)[C@@H]1O</td>\n",
              "      <td>245.24</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1944</th>\n",
              "      <td>CO[C@H]1O[C@H](CO)[C@H](O)[C@H](O)[C@@H]1O</td>\n",
              "      <td>245.22</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1918</th>\n",
              "      <td>CSCC[C@@H](N)C(=O)O</td>\n",
              "      <td>194.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>276</th>\n",
              "      <td>CSCC[C@@H](N)C(=O)O</td>\n",
              "      <td>200.74</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>N[C@@H](CC(=O)O)C(=O)O</td>\n",
              "      <td>218.32</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1469</th>\n",
              "      <td>N[C@@H](CC(=O)O)C(=O)O</td>\n",
              "      <td>218.34</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>135</th>\n",
              "      <td>N[C@@H](CCC(=O)O)C(=O)O</td>\n",
              "      <td>229.92</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>475</th>\n",
              "      <td>N[C@@H](CCC(=O)O)C(=O)O</td>\n",
              "      <td>229.92</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>871</th>\n",
              "      <td>N[C@@H](CO)C(=O)O</td>\n",
              "      <td>229.72</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1730</th>\n",
              "      <td>N[C@@H](CO)C(=O)O</td>\n",
              "      <td>233.94</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1109</th>\n",
              "      <td>N[C@@H](Cc1ccccc1)C(=O)O</td>\n",
              "      <td>197.56</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>313</th>\n",
              "      <td>N[C@@H](Cc1ccccc1)C(=O)O</td>\n",
              "      <td>197.53</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2540</th>\n",
              "      <td>N[C@H]1CCCC[C@@H]1N</td>\n",
              "      <td>218.19</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2194</th>\n",
              "      <td>N[C@H]1CCCC[C@@H]1N</td>\n",
              "      <td>218.19</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1361</th>\n",
              "      <td>O=C(/C=C/c1ccccc1)c1ccccc1</td>\n",
              "      <td>146.51</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>921</th>\n",
              "      <td>O=C(/C=C/c1ccccc1)c1ccccc1</td>\n",
              "      <td>146.51</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>643</th>\n",
              "      <td>O=C(O)[C@H]1CCCN1</td>\n",
              "      <td>213.07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>254</th>\n",
              "      <td>O=C(O)[C@H]1CCCN1</td>\n",
              "      <td>219.57</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1542</th>\n",
              "      <td>O=C(O)[C@H]1CCCN1</td>\n",
              "      <td>219.57</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>914</th>\n",
              "      <td>O=C(O)c1ccccc1</td>\n",
              "      <td>186.42</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2011</th>\n",
              "      <td>O=C(O)c1ccccc1</td>\n",
              "      <td>186.43</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2231</th>\n",
              "      <td>O=C1C=CCCC1</td>\n",
              "      <td>188.30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>541</th>\n",
              "      <td>O=C1C=CCCC1</td>\n",
              "      <td>188.30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1183</th>\n",
              "      <td>O=C1OC(=O)c2ncccc21</td>\n",
              "      <td>182.41</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2715</th>\n",
              "      <td>O=C1OC(=O)c2ncccc21</td>\n",
              "      <td>182.41</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>594</th>\n",
              "      <td>OC[C@H]1O[C@H](O)[C@H](O)[C@H](O)[C@@H]1O</td>\n",
              "      <td>231.96</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2381</th>\n",
              "      <td>OC[C@H]1O[C@H](O)[C@H](O)[C@H](O)[C@@H]1O</td>\n",
              "      <td>231.93</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1294</th>\n",
              "      <td>c1cnc2[nH]ccc2c1</td>\n",
              "      <td>169.53</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1732</th>\n",
              "      <td>c1cnc2[nH]ccc2c1</td>\n",
              "      <td>169.53</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                          SMILES  Energygap\n",
              "218                                  C1=CCCCCCC1     214.10\n",
              "2125                                 C1=CCCCCCC1     214.10\n",
              "1901                     C1CC[C@H]2CCCC[C@@H]2C1     251.14\n",
              "552                      C1CC[C@H]2CCCC[C@@H]2C1     251.14\n",
              "554                                 C=CN1CCCC1=O     190.83\n",
              "1808                                C=CN1CCCC1=O     190.84\n",
              "2203                                 C=Cc1ccccc1     168.47\n",
              "1354                                 C=Cc1ccccc1     168.48\n",
              "2341                                 C=Cc1ccccc1     168.49\n",
              "1265                                 C=Cc1ccncc1     174.83\n",
              "265                                  C=Cc1ccncc1     174.85\n",
              "2400              CC(C)(C)OC(=O)N1CCC(C(=O)O)CC1     208.68\n",
              "1621              CC(C)(C)OC(=O)N1CCC(C(=O)O)CC1     207.40\n",
              "1184                       CC(C)C[C@@H](N)C(=O)O     224.37\n",
              "1759                       CC(C)C[C@@H](N)C(=O)O     224.48\n",
              "2324                                      CC(C)I     192.15\n",
              "85                                        CC(C)I     192.15\n",
              "2885                        CC(C)[C@@H](N)C(=O)O     225.26\n",
              "2408                        CC(C)[C@@H](N)C(=O)O     225.27\n",
              "21                          CC(C)[C@@H](N)C(=O)O     225.28\n",
              "1856                            CC(C)[C@@H](N)CO     233.36\n",
              "2813                            CC(C)[C@@H](N)CO     233.36\n",
              "1841             CC(C)[C@@H]1CC[C@@H](C)C[C@H]1O     239.02\n",
              "1448             CC(C)[C@@H]1CC[C@@H](C)C[C@H]1O     239.01\n",
              "1881               CC1(C)[C@H]2CC=C(CO)[C@@H]1C2     204.96\n",
              "819                CC1(C)[C@H]2CC=C(CO)[C@@H]1C2     204.96\n",
              "1635                  CCCCCCCC/C=C/CCCCCCCC(=O)O     214.11\n",
              "2310                  CCCCCCCC/C=C/CCCCCCCC(=O)O     216.25\n",
              "262                                           CI     195.30\n",
              "853                                           CI     195.29\n",
              "2061  CO[C@H]1O[C@H](CO)[C@H](O)[C@H](O)[C@@H]1O     245.24\n",
              "1944  CO[C@H]1O[C@H](CO)[C@H](O)[C@H](O)[C@@H]1O     245.22\n",
              "1918                         CSCC[C@@H](N)C(=O)O     194.00\n",
              "276                          CSCC[C@@H](N)C(=O)O     200.74\n",
              "11                        N[C@@H](CC(=O)O)C(=O)O     218.32\n",
              "1469                      N[C@@H](CC(=O)O)C(=O)O     218.34\n",
              "135                      N[C@@H](CCC(=O)O)C(=O)O     229.92\n",
              "475                      N[C@@H](CCC(=O)O)C(=O)O     229.92\n",
              "871                            N[C@@H](CO)C(=O)O     229.72\n",
              "1730                           N[C@@H](CO)C(=O)O     233.94\n",
              "1109                    N[C@@H](Cc1ccccc1)C(=O)O     197.56\n",
              "313                     N[C@@H](Cc1ccccc1)C(=O)O     197.53\n",
              "2540                         N[C@H]1CCCC[C@@H]1N     218.19\n",
              "2194                         N[C@H]1CCCC[C@@H]1N     218.19\n",
              "1361                  O=C(/C=C/c1ccccc1)c1ccccc1     146.51\n",
              "921                   O=C(/C=C/c1ccccc1)c1ccccc1     146.51\n",
              "643                            O=C(O)[C@H]1CCCN1     213.07\n",
              "254                            O=C(O)[C@H]1CCCN1     219.57\n",
              "1542                           O=C(O)[C@H]1CCCN1     219.57\n",
              "914                               O=C(O)c1ccccc1     186.42\n",
              "2011                              O=C(O)c1ccccc1     186.43\n",
              "2231                                 O=C1C=CCCC1     188.30\n",
              "541                                  O=C1C=CCCC1     188.30\n",
              "1183                         O=C1OC(=O)c2ncccc21     182.41\n",
              "2715                         O=C1OC(=O)c2ncccc21     182.41\n",
              "594    OC[C@H]1O[C@H](O)[C@H](O)[C@H](O)[C@@H]1O     231.96\n",
              "2381   OC[C@H]1O[C@H](O)[C@H](O)[C@H](O)[C@@H]1O     231.93\n",
              "1294                            c1cnc2[nH]ccc2c1     169.53\n",
              "1732                            c1cnc2[nH]ccc2c1     169.53"
            ]
          },
          "execution_count": 76,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# Create a list for duplicate smiles\n",
        "dataset[dataset['SMILES'].isin(duplicates_smiles)].sort_values(by=['SMILES'])"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "V0_GH3m3dlvA",
      "metadata": {
        "id": "V0_GH3m3dlvA"
      },
      "source": [
        "## 2.  Drop duplicate values"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 77,
      "id": "TdrRU3Y0dlvA",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TdrRU3Y0dlvA",
        "outputId": "3933b967-6ae1-4c8f-a186-fe6509d6c0a6"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "2873"
            ]
          },
          "execution_count": 77,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "dataset_new = dataset.drop_duplicates(subset=['SMILES'])\n",
        "len(dataset_new)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 78,
      "id": "F1OXOWPfdlvB",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "id": "F1OXOWPfdlvB",
        "outputId": "12870e83-1bd5-48ff-e1d6-e894f94e1e28"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>SMILES</th>\n",
              "      <th>Energygap</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Cc1ccc(C(F)(F)F)cc1</td>\n",
              "      <td>197.75</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>O=C(O)CCCCl</td>\n",
              "      <td>247.49</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>CC(C)(Oc1ccc(CCNC(=O)c2ccc(Cl)cc2)cc1)C(=O)O</td>\n",
              "      <td>164.71</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Nc1ccc(Cl)c(Cl)c1</td>\n",
              "      <td>169.03</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>CC(C)=CCC[C@@H](C)CCO</td>\n",
              "      <td>209.57</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2899</th>\n",
              "      <td>c1ccc(P(CCP(c2ccccc2)c2ccccc2)c2ccccc2)cc1</td>\n",
              "      <td>168.65</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2900</th>\n",
              "      <td>Brc1cccc2sccc12</td>\n",
              "      <td>162.93</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2901</th>\n",
              "      <td>CCOC(=O)N1c2ccccc2C=C[C@@H]1OCC</td>\n",
              "      <td>165.10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2902</th>\n",
              "      <td>c1ccc2sccc2c1</td>\n",
              "      <td>167.96</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2903</th>\n",
              "      <td>Cc1cccc(C)c1O</td>\n",
              "      <td>188.37</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2873 rows × 2 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "                                            SMILES  Energygap\n",
              "0                              Cc1ccc(C(F)(F)F)cc1     197.75\n",
              "1                                      O=C(O)CCCCl     247.49\n",
              "2     CC(C)(Oc1ccc(CCNC(=O)c2ccc(Cl)cc2)cc1)C(=O)O     164.71\n",
              "3                                Nc1ccc(Cl)c(Cl)c1     169.03\n",
              "4                            CC(C)=CCC[C@@H](C)CCO     209.57\n",
              "...                                            ...        ...\n",
              "2899    c1ccc(P(CCP(c2ccccc2)c2ccccc2)c2ccccc2)cc1     168.65\n",
              "2900                               Brc1cccc2sccc12     162.93\n",
              "2901               CCOC(=O)N1c2ccccc2C=C[C@@H]1OCC     165.10\n",
              "2902                                 c1ccc2sccc2c1     167.96\n",
              "2903                                 Cc1cccc(C)c1O     188.37\n",
              "\n",
              "[2873 rows x 2 columns]"
            ]
          },
          "execution_count": 78,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "dataset_new"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 79,
      "id": "ALH0makabOon",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 387
        },
        "id": "ALH0makabOon",
        "outputId": "6eab0be5-a528-4909-c7a3-fc6b6c6efbab"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<seaborn.axisgrid.FacetGrid at 0x233cc1f87d0>"
            ]
          },
          "execution_count": 79,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 500x500 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "sn.displot(dataset_new.Energygap)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 80,
      "id": "Oos7RJuYbkAR",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Oos7RJuYbkAR",
        "outputId": "cc7fe466-85d5-4ada-eb39-43c7a1bdd54e"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "count   2873.00\n",
              "mean     190.64\n",
              "std       30.66\n",
              "min       78.99\n",
              "25%      169.40\n",
              "50%      184.80\n",
              "75%      211.16\n",
              "max      309.98\n",
              "Name: Energygap, dtype: float64"
            ]
          },
          "execution_count": 80,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "dataset_new.Energygap.describe()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 81,
      "id": "BdUBRjxncDMS",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BdUBRjxncDMS",
        "outputId": "df5995ae-029e-468b-d244-8b60eb1ead99"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "184.7951254"
            ]
          },
          "execution_count": 81,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "np.median(dataset_new.Energygap)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 82,
      "id": "c6B-Oi1icRNL",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 297
        },
        "id": "c6B-Oi1icRNL",
        "outputId": "c3309ca9-6917-410e-e584-9e36875da59f"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<Axes: >"
            ]
          },
          "execution_count": 82,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "sn.boxplot(dataset_new.Energygap)"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "xBg1zvledlvB",
      "metadata": {
        "id": "xBg1zvledlvB"
      },
      "source": [
        "## Calculate descriptors using RDkit"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "iG6kBPvSdlvB",
      "metadata": {
        "id": "iG6kBPvSdlvB"
      },
      "source": [
        "###  General molecular descriptors-about 200 molecular descriptors"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 83,
      "id": "HPr3DWj5dlvB",
      "metadata": {
        "id": "HPr3DWj5dlvB"
      },
      "outputs": [],
      "source": [
        "def RDkit_descriptors(smiles):\n",
        "    mols = [Chem.MolFromSmiles(i) for i in smiles] \n",
        "    calc = MoleculeDescriptors.MolecularDescriptorCalculator([x[0] \n",
        "                                    for x in Descriptors._descList])\n",
        "    desc_names = calc.GetDescriptorNames()\n",
        "    \n",
        "    Mol_descriptors =[]\n",
        "    for mol in mols:\n",
        "        # add hydrogens to molecules\n",
        "        mol=Chem.AddHs(mol)\n",
        "        # Calculate all 200 descriptors for each molecule\n",
        "        descriptors = calc.CalcDescriptors(mol)\n",
        "        Mol_descriptors.append(descriptors)\n",
        "    return Mol_descriptors,desc_names \n",
        "\n",
        "# Function call\n",
        "Mol_descriptors,desc_names = RDkit_descriptors(dataset_new['SMILES'])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 84,
      "id": "2hFGOZwCdlvB",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 488
        },
        "id": "2hFGOZwCdlvB",
        "outputId": "8ad3338c-1f6b-42fc-bfb8-843b10d33a57",
        "scrolled": true
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>MaxEStateIndex</th>\n",
              "      <th>MinEStateIndex</th>\n",
              "      <th>MaxAbsEStateIndex</th>\n",
              "      <th>MinAbsEStateIndex</th>\n",
              "      <th>qed</th>\n",
              "      <th>MolWt</th>\n",
              "      <th>HeavyAtomMolWt</th>\n",
              "      <th>ExactMolWt</th>\n",
              "      <th>NumValenceElectrons</th>\n",
              "      <th>NumRadicalElectrons</th>\n",
              "      <th>...</th>\n",
              "      <th>fr_sulfide</th>\n",
              "      <th>fr_sulfonamd</th>\n",
              "      <th>fr_sulfone</th>\n",
              "      <th>fr_term_acetylene</th>\n",
              "      <th>fr_tetrazole</th>\n",
              "      <th>fr_thiazole</th>\n",
              "      <th>fr_thiocyan</th>\n",
              "      <th>fr_thiophene</th>\n",
              "      <th>fr_unbrch_alkane</th>\n",
              "      <th>fr_urea</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>12.55</td>\n",
              "      <td>-5.08</td>\n",
              "      <td>12.55</td>\n",
              "      <td>1.01</td>\n",
              "      <td>0.55</td>\n",
              "      <td>160.14</td>\n",
              "      <td>153.08</td>\n",
              "      <td>160.05</td>\n",
              "      <td>60</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>10.68</td>\n",
              "      <td>-3.33</td>\n",
              "      <td>10.68</td>\n",
              "      <td>1.84</td>\n",
              "      <td>0.57</td>\n",
              "      <td>122.55</td>\n",
              "      <td>115.49</td>\n",
              "      <td>122.01</td>\n",
              "      <td>42</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>13.05</td>\n",
              "      <td>-4.11</td>\n",
              "      <td>13.05</td>\n",
              "      <td>0.72</td>\n",
              "      <td>0.79</td>\n",
              "      <td>361.82</td>\n",
              "      <td>341.67</td>\n",
              "      <td>361.11</td>\n",
              "      <td>132</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>7.40</td>\n",
              "      <td>-0.45</td>\n",
              "      <td>7.40</td>\n",
              "      <td>0.07</td>\n",
              "      <td>0.58</td>\n",
              "      <td>162.02</td>\n",
              "      <td>156.98</td>\n",
              "      <td>160.98</td>\n",
              "      <td>48</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>8.10</td>\n",
              "      <td>-4.48</td>\n",
              "      <td>8.10</td>\n",
              "      <td>1.89</td>\n",
              "      <td>0.61</td>\n",
              "      <td>156.27</td>\n",
              "      <td>136.11</td>\n",
              "      <td>156.15</td>\n",
              "      <td>66</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2868</th>\n",
              "      <td>9.51</td>\n",
              "      <td>-3.87</td>\n",
              "      <td>9.51</td>\n",
              "      <td>0.97</td>\n",
              "      <td>0.37</td>\n",
              "      <td>398.43</td>\n",
              "      <td>374.23</td>\n",
              "      <td>398.14</td>\n",
              "      <td>138</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2869</th>\n",
              "      <td>7.65</td>\n",
              "      <td>-0.17</td>\n",
              "      <td>7.65</td>\n",
              "      <td>0.02</td>\n",
              "      <td>0.63</td>\n",
              "      <td>213.10</td>\n",
              "      <td>208.06</td>\n",
              "      <td>211.93</td>\n",
              "      <td>50</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2870</th>\n",
              "      <td>13.02</td>\n",
              "      <td>-3.81</td>\n",
              "      <td>13.02</td>\n",
              "      <td>0.33</td>\n",
              "      <td>0.82</td>\n",
              "      <td>247.29</td>\n",
              "      <td>230.16</td>\n",
              "      <td>247.12</td>\n",
              "      <td>96</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2871</th>\n",
              "      <td>7.59</td>\n",
              "      <td>-0.35</td>\n",
              "      <td>7.59</td>\n",
              "      <td>0.03</td>\n",
              "      <td>0.52</td>\n",
              "      <td>134.20</td>\n",
              "      <td>128.15</td>\n",
              "      <td>134.02</td>\n",
              "      <td>44</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2872</th>\n",
              "      <td>7.52</td>\n",
              "      <td>-2.92</td>\n",
              "      <td>7.52</td>\n",
              "      <td>0.82</td>\n",
              "      <td>0.56</td>\n",
              "      <td>122.17</td>\n",
              "      <td>112.09</td>\n",
              "      <td>122.07</td>\n",
              "      <td>48</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2873 rows × 208 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      MaxEStateIndex  MinEStateIndex  MaxAbsEStateIndex  MinAbsEStateIndex  \\\n",
              "0              12.55           -5.08              12.55               1.01   \n",
              "1              10.68           -3.33              10.68               1.84   \n",
              "2              13.05           -4.11              13.05               0.72   \n",
              "3               7.40           -0.45               7.40               0.07   \n",
              "4               8.10           -4.48               8.10               1.89   \n",
              "...              ...             ...                ...                ...   \n",
              "2868            9.51           -3.87               9.51               0.97   \n",
              "2869            7.65           -0.17               7.65               0.02   \n",
              "2870           13.02           -3.81              13.02               0.33   \n",
              "2871            7.59           -0.35               7.59               0.03   \n",
              "2872            7.52           -2.92               7.52               0.82   \n",
              "\n",
              "      qed  MolWt  HeavyAtomMolWt  ExactMolWt  NumValenceElectrons  \\\n",
              "0    0.55 160.14          153.08      160.05                   60   \n",
              "1    0.57 122.55          115.49      122.01                   42   \n",
              "2    0.79 361.82          341.67      361.11                  132   \n",
              "3    0.58 162.02          156.98      160.98                   48   \n",
              "4    0.61 156.27          136.11      156.15                   66   \n",
              "...   ...    ...             ...         ...                  ...   \n",
              "2868 0.37 398.43          374.23      398.14                  138   \n",
              "2869 0.63 213.10          208.06      211.93                   50   \n",
              "2870 0.82 247.29          230.16      247.12                   96   \n",
              "2871 0.52 134.20          128.15      134.02                   44   \n",
              "2872 0.56 122.17          112.09      122.07                   48   \n",
              "\n",
              "      NumRadicalElectrons  ...  fr_sulfide  fr_sulfonamd  fr_sulfone  \\\n",
              "0                       0  ...           0             0           0   \n",
              "1                       0  ...           0             0           0   \n",
              "2                       0  ...           0             0           0   \n",
              "3                       0  ...           0             0           0   \n",
              "4                       0  ...           0             0           0   \n",
              "...                   ...  ...         ...           ...         ...   \n",
              "2868                    0  ...           0             0           0   \n",
              "2869                    0  ...           0             0           0   \n",
              "2870                    0  ...           0             0           0   \n",
              "2871                    0  ...           0             0           0   \n",
              "2872                    0  ...           0             0           0   \n",
              "\n",
              "      fr_term_acetylene  fr_tetrazole  fr_thiazole  fr_thiocyan  fr_thiophene  \\\n",
              "0                     0             0            0            0             0   \n",
              "1                     0             0            0            0             0   \n",
              "2                     0             0            0            0             0   \n",
              "3                     0             0            0            0             0   \n",
              "4                     0             0            0            0             0   \n",
              "...                 ...           ...          ...          ...           ...   \n",
              "2868                  0             0            0            0             0   \n",
              "2869                  0             0            0            0             1   \n",
              "2870                  0             0            0            0             0   \n",
              "2871                  0             0            0            0             1   \n",
              "2872                  0             0            0            0             0   \n",
              "\n",
              "      fr_unbrch_alkane  fr_urea  \n",
              "0                    0        0  \n",
              "1                    0        0  \n",
              "2                    0        0  \n",
              "3                    0        0  \n",
              "4                    0        0  \n",
              "...                ...      ...  \n",
              "2868                 0        0  \n",
              "2869                 0        0  \n",
              "2870                 0        0  \n",
              "2871                 0        0  \n",
              "2872                 0        0  \n",
              "\n",
              "[2873 rows x 208 columns]"
            ]
          },
          "execution_count": 84,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df_with_200_descriptors = pd.DataFrame(Mol_descriptors,columns=desc_names)\n",
        "df_with_200_descriptors"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "bgEs5YWFdlvB",
      "metadata": {
        "id": "bgEs5YWFdlvB"
      },
      "source": [
        "# Part II - HOMO-LUMO Energy gap prediction of organic compounds using machine learning"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "Oyu9r5NvdlvC",
      "metadata": {
        "id": "Oyu9r5NvdlvC"
      },
      "source": [
        "## 1. Remove highly correlated features"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 85,
      "id": "vn9yBv7UdlvC",
      "metadata": {
        "id": "vn9yBv7UdlvC"
      },
      "outputs": [],
      "source": [
        "def remove_correlated_features(descriptors):\n",
        "    # Calculate correlation\n",
        "    correlated_matrix = descriptors.corr().abs()\n",
        "\n",
        "    # Upper triangle of correlation matrix\n",
        "    upper_triangle = correlated_matrix.where(np.triu(np.ones(correlated_matrix.shape), k=1).astype(bool))\n",
        "\n",
        "    # Identify columns that have above 0.9 values of correlation\n",
        "    to_drop = [column for column in upper_triangle.columns if any(upper_triangle[column] >= 0.9)]\n",
        "    print(to_drop)\n",
        "    descriptors_correlated_dropped = descriptors.drop(columns=to_drop, axis=1)\n",
        "    return descriptors_correlated_dropped    "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 86,
      "id": "hVcUJdhXdlvC",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 506
        },
        "id": "hVcUJdhXdlvC",
        "outputId": "4f25043b-f2dd-4813-e732-bf936922f566"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['MaxAbsEStateIndex', 'HeavyAtomMolWt', 'ExactMolWt', 'MaxAbsPartialCharge', 'MinAbsPartialCharge', 'FpDensityMorgan2', 'FpDensityMorgan3', 'BCUT2D_LOGPHI', 'Chi0n', 'Chi0v', 'Chi1', 'Chi1n', 'Chi1v', 'Chi2n', 'Chi3n', 'Chi4n', 'Chi4v', 'LabuteASA', 'SlogP_VSA5', 'SlogP_VSA6', 'EState_VSA1', 'VSA_EState1', 'HeavyAtomCount', 'NOCount', 'NumHAcceptors', 'NumHDonors', 'MolMR', 'fr_COO2', 'fr_C_O', 'fr_Nhpyrrole', 'fr_benzene', 'fr_nitro_arom', 'fr_phenol', 'fr_phenol_noOrthoHbond']\n"
          ]
        },
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>MaxEStateIndex</th>\n",
              "      <th>MinEStateIndex</th>\n",
              "      <th>MinAbsEStateIndex</th>\n",
              "      <th>qed</th>\n",
              "      <th>MolWt</th>\n",
              "      <th>NumValenceElectrons</th>\n",
              "      <th>NumRadicalElectrons</th>\n",
              "      <th>MaxPartialCharge</th>\n",
              "      <th>MinPartialCharge</th>\n",
              "      <th>FpDensityMorgan1</th>\n",
              "      <th>...</th>\n",
              "      <th>fr_sulfide</th>\n",
              "      <th>fr_sulfonamd</th>\n",
              "      <th>fr_sulfone</th>\n",
              "      <th>fr_term_acetylene</th>\n",
              "      <th>fr_tetrazole</th>\n",
              "      <th>fr_thiazole</th>\n",
              "      <th>fr_thiocyan</th>\n",
              "      <th>fr_thiophene</th>\n",
              "      <th>fr_unbrch_alkane</th>\n",
              "      <th>fr_urea</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>12.55</td>\n",
              "      <td>-5.08</td>\n",
              "      <td>1.01</td>\n",
              "      <td>0.55</td>\n",
              "      <td>160.14</td>\n",
              "      <td>60</td>\n",
              "      <td>0</td>\n",
              "      <td>0.42</td>\n",
              "      <td>-0.17</td>\n",
              "      <td>1.27</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>10.68</td>\n",
              "      <td>-3.33</td>\n",
              "      <td>1.84</td>\n",
              "      <td>0.57</td>\n",
              "      <td>122.55</td>\n",
              "      <td>42</td>\n",
              "      <td>0</td>\n",
              "      <td>0.30</td>\n",
              "      <td>-0.48</td>\n",
              "      <td>2.14</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>13.05</td>\n",
              "      <td>-4.11</td>\n",
              "      <td>0.72</td>\n",
              "      <td>0.79</td>\n",
              "      <td>361.82</td>\n",
              "      <td>132</td>\n",
              "      <td>0</td>\n",
              "      <td>0.35</td>\n",
              "      <td>-0.48</td>\n",
              "      <td>1.32</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>7.40</td>\n",
              "      <td>-0.45</td>\n",
              "      <td>0.07</td>\n",
              "      <td>0.58</td>\n",
              "      <td>162.02</td>\n",
              "      <td>48</td>\n",
              "      <td>0</td>\n",
              "      <td>0.16</td>\n",
              "      <td>-0.40</td>\n",
              "      <td>1.44</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>8.10</td>\n",
              "      <td>-4.48</td>\n",
              "      <td>1.89</td>\n",
              "      <td>0.61</td>\n",
              "      <td>156.27</td>\n",
              "      <td>66</td>\n",
              "      <td>0</td>\n",
              "      <td>0.21</td>\n",
              "      <td>-0.40</td>\n",
              "      <td>1.91</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2868</th>\n",
              "      <td>9.51</td>\n",
              "      <td>-3.87</td>\n",
              "      <td>0.97</td>\n",
              "      <td>0.37</td>\n",
              "      <td>398.43</td>\n",
              "      <td>138</td>\n",
              "      <td>0</td>\n",
              "      <td>0.06</td>\n",
              "      <td>-0.06</td>\n",
              "      <td>0.43</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2869</th>\n",
              "      <td>7.65</td>\n",
              "      <td>-0.17</td>\n",
              "      <td>0.02</td>\n",
              "      <td>0.63</td>\n",
              "      <td>213.10</td>\n",
              "      <td>50</td>\n",
              "      <td>0</td>\n",
              "      <td>0.07</td>\n",
              "      <td>-0.14</td>\n",
              "      <td>1.40</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2870</th>\n",
              "      <td>13.02</td>\n",
              "      <td>-3.81</td>\n",
              "      <td>0.33</td>\n",
              "      <td>0.82</td>\n",
              "      <td>247.29</td>\n",
              "      <td>96</td>\n",
              "      <td>0</td>\n",
              "      <td>0.42</td>\n",
              "      <td>-0.45</td>\n",
              "      <td>1.56</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2871</th>\n",
              "      <td>7.59</td>\n",
              "      <td>-0.35</td>\n",
              "      <td>0.03</td>\n",
              "      <td>0.52</td>\n",
              "      <td>134.20</td>\n",
              "      <td>44</td>\n",
              "      <td>0</td>\n",
              "      <td>0.07</td>\n",
              "      <td>-0.14</td>\n",
              "      <td>1.22</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2872</th>\n",
              "      <td>7.52</td>\n",
              "      <td>-2.92</td>\n",
              "      <td>0.82</td>\n",
              "      <td>0.56</td>\n",
              "      <td>122.17</td>\n",
              "      <td>48</td>\n",
              "      <td>0</td>\n",
              "      <td>0.29</td>\n",
              "      <td>-0.51</td>\n",
              "      <td>1.56</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2873 rows × 174 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      MaxEStateIndex  MinEStateIndex  MinAbsEStateIndex  qed  MolWt  \\\n",
              "0              12.55           -5.08               1.01 0.55 160.14   \n",
              "1              10.68           -3.33               1.84 0.57 122.55   \n",
              "2              13.05           -4.11               0.72 0.79 361.82   \n",
              "3               7.40           -0.45               0.07 0.58 162.02   \n",
              "4               8.10           -4.48               1.89 0.61 156.27   \n",
              "...              ...             ...                ...  ...    ...   \n",
              "2868            9.51           -3.87               0.97 0.37 398.43   \n",
              "2869            7.65           -0.17               0.02 0.63 213.10   \n",
              "2870           13.02           -3.81               0.33 0.82 247.29   \n",
              "2871            7.59           -0.35               0.03 0.52 134.20   \n",
              "2872            7.52           -2.92               0.82 0.56 122.17   \n",
              "\n",
              "      NumValenceElectrons  NumRadicalElectrons  MaxPartialCharge  \\\n",
              "0                      60                    0              0.42   \n",
              "1                      42                    0              0.30   \n",
              "2                     132                    0              0.35   \n",
              "3                      48                    0              0.16   \n",
              "4                      66                    0              0.21   \n",
              "...                   ...                  ...               ...   \n",
              "2868                  138                    0              0.06   \n",
              "2869                   50                    0              0.07   \n",
              "2870                   96                    0              0.42   \n",
              "2871                   44                    0              0.07   \n",
              "2872                   48                    0              0.29   \n",
              "\n",
              "      MinPartialCharge  FpDensityMorgan1  ...  fr_sulfide  fr_sulfonamd  \\\n",
              "0                -0.17              1.27  ...           0             0   \n",
              "1                -0.48              2.14  ...           0             0   \n",
              "2                -0.48              1.32  ...           0             0   \n",
              "3                -0.40              1.44  ...           0             0   \n",
              "4                -0.40              1.91  ...           0             0   \n",
              "...                ...               ...  ...         ...           ...   \n",
              "2868             -0.06              0.43  ...           0             0   \n",
              "2869             -0.14              1.40  ...           0             0   \n",
              "2870             -0.45              1.56  ...           0             0   \n",
              "2871             -0.14              1.22  ...           0             0   \n",
              "2872             -0.51              1.56  ...           0             0   \n",
              "\n",
              "      fr_sulfone  fr_term_acetylene  fr_tetrazole  fr_thiazole  fr_thiocyan  \\\n",
              "0              0                  0             0            0            0   \n",
              "1              0                  0             0            0            0   \n",
              "2              0                  0             0            0            0   \n",
              "3              0                  0             0            0            0   \n",
              "4              0                  0             0            0            0   \n",
              "...          ...                ...           ...          ...          ...   \n",
              "2868           0                  0             0            0            0   \n",
              "2869           0                  0             0            0            0   \n",
              "2870           0                  0             0            0            0   \n",
              "2871           0                  0             0            0            0   \n",
              "2872           0                  0             0            0            0   \n",
              "\n",
              "      fr_thiophene  fr_unbrch_alkane  fr_urea  \n",
              "0                0                 0        0  \n",
              "1                0                 0        0  \n",
              "2                0                 0        0  \n",
              "3                0                 0        0  \n",
              "4                0                 0        0  \n",
              "...            ...               ...      ...  \n",
              "2868             0                 0        0  \n",
              "2869             1                 0        0  \n",
              "2870             0                 0        0  \n",
              "2871             1                 0        0  \n",
              "2872             0                 0        0  \n",
              "\n",
              "[2873 rows x 174 columns]"
            ]
          },
          "execution_count": 86,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "descriptors_new = remove_correlated_features(df_with_200_descriptors)\n",
        "descriptors_new"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 87,
      "id": "NxvZDIYOdlvC",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 488
        },
        "id": "NxvZDIYOdlvC",
        "outputId": "4922a237-f963-4213-efb5-cea3f0f2cef1"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>MaxEStateIndex</th>\n",
              "      <th>MinEStateIndex</th>\n",
              "      <th>MinAbsEStateIndex</th>\n",
              "      <th>MolWt</th>\n",
              "      <th>NumValenceElectrons</th>\n",
              "      <th>NumRadicalElectrons</th>\n",
              "      <th>FpDensityMorgan1</th>\n",
              "      <th>BCUT2D_MWHI</th>\n",
              "      <th>BCUT2D_MWLOW</th>\n",
              "      <th>BCUT2D_CHGHI</th>\n",
              "      <th>...</th>\n",
              "      <th>fr_amide</th>\n",
              "      <th>fr_aniline</th>\n",
              "      <th>fr_bicyclic</th>\n",
              "      <th>fr_ester</th>\n",
              "      <th>fr_ether</th>\n",
              "      <th>fr_halogen</th>\n",
              "      <th>fr_ketone</th>\n",
              "      <th>fr_ketone_Topliss</th>\n",
              "      <th>fr_methoxy</th>\n",
              "      <th>fr_para_hydroxylation</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>12.55</td>\n",
              "      <td>-5.08</td>\n",
              "      <td>1.01</td>\n",
              "      <td>160.14</td>\n",
              "      <td>60</td>\n",
              "      <td>0</td>\n",
              "      <td>1.27</td>\n",
              "      <td>11.63</td>\n",
              "      <td>-0.24</td>\n",
              "      <td>0.00</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>10.68</td>\n",
              "      <td>-3.33</td>\n",
              "      <td>1.84</td>\n",
              "      <td>122.55</td>\n",
              "      <td>42</td>\n",
              "      <td>0</td>\n",
              "      <td>2.14</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.19</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>13.05</td>\n",
              "      <td>-4.11</td>\n",
              "      <td>0.72</td>\n",
              "      <td>361.82</td>\n",
              "      <td>132</td>\n",
              "      <td>0</td>\n",
              "      <td>1.32</td>\n",
              "      <td>10.86</td>\n",
              "      <td>-0.24</td>\n",
              "      <td>0.00</td>\n",
              "      <td>...</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>7.40</td>\n",
              "      <td>-0.45</td>\n",
              "      <td>0.07</td>\n",
              "      <td>162.02</td>\n",
              "      <td>48</td>\n",
              "      <td>0</td>\n",
              "      <td>1.44</td>\n",
              "      <td>12.64</td>\n",
              "      <td>-0.14</td>\n",
              "      <td>0.51</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>8.10</td>\n",
              "      <td>-4.48</td>\n",
              "      <td>1.89</td>\n",
              "      <td>156.27</td>\n",
              "      <td>66</td>\n",
              "      <td>0</td>\n",
              "      <td>1.91</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.25</td>\n",
              "      <td>-0.04</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2868</th>\n",
              "      <td>9.51</td>\n",
              "      <td>-3.87</td>\n",
              "      <td>0.97</td>\n",
              "      <td>398.43</td>\n",
              "      <td>138</td>\n",
              "      <td>0</td>\n",
              "      <td>0.43</td>\n",
              "      <td>10.46</td>\n",
              "      <td>-0.18</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2869</th>\n",
              "      <td>7.65</td>\n",
              "      <td>-0.17</td>\n",
              "      <td>0.02</td>\n",
              "      <td>213.10</td>\n",
              "      <td>50</td>\n",
              "      <td>0</td>\n",
              "      <td>1.40</td>\n",
              "      <td>12.60</td>\n",
              "      <td>-0.09</td>\n",
              "      <td>0.55</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2870</th>\n",
              "      <td>13.02</td>\n",
              "      <td>-3.81</td>\n",
              "      <td>0.33</td>\n",
              "      <td>247.29</td>\n",
              "      <td>96</td>\n",
              "      <td>0</td>\n",
              "      <td>1.56</td>\n",
              "      <td>10.13</td>\n",
              "      <td>-0.25</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>...</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2871</th>\n",
              "      <td>7.59</td>\n",
              "      <td>-0.35</td>\n",
              "      <td>0.03</td>\n",
              "      <td>134.20</td>\n",
              "      <td>44</td>\n",
              "      <td>0</td>\n",
              "      <td>1.22</td>\n",
              "      <td>11.25</td>\n",
              "      <td>-0.09</td>\n",
              "      <td>0.51</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2872</th>\n",
              "      <td>7.52</td>\n",
              "      <td>-2.92</td>\n",
              "      <td>0.82</td>\n",
              "      <td>122.17</td>\n",
              "      <td>48</td>\n",
              "      <td>0</td>\n",
              "      <td>1.56</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.25</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2873 rows × 106 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      MaxEStateIndex  MinEStateIndex  MinAbsEStateIndex  MolWt  \\\n",
              "0              12.55           -5.08               1.01 160.14   \n",
              "1              10.68           -3.33               1.84 122.55   \n",
              "2              13.05           -4.11               0.72 361.82   \n",
              "3               7.40           -0.45               0.07 162.02   \n",
              "4               8.10           -4.48               1.89 156.27   \n",
              "...              ...             ...                ...    ...   \n",
              "2868            9.51           -3.87               0.97 398.43   \n",
              "2869            7.65           -0.17               0.02 213.10   \n",
              "2870           13.02           -3.81               0.33 247.29   \n",
              "2871            7.59           -0.35               0.03 134.20   \n",
              "2872            7.52           -2.92               0.82 122.17   \n",
              "\n",
              "      NumValenceElectrons  NumRadicalElectrons  FpDensityMorgan1  BCUT2D_MWHI  \\\n",
              "0                      60                    0              1.27        11.63   \n",
              "1                      42                    0              2.14         0.00   \n",
              "2                     132                    0              1.32        10.86   \n",
              "3                      48                    0              1.44        12.64   \n",
              "4                      66                    0              1.91        -0.00   \n",
              "...                   ...                  ...               ...          ...   \n",
              "2868                  138                    0              0.43        10.46   \n",
              "2869                   50                    0              1.40        12.60   \n",
              "2870                   96                    0              1.56        10.13   \n",
              "2871                   44                    0              1.22        11.25   \n",
              "2872                   48                    0              1.56         0.00   \n",
              "\n",
              "      BCUT2D_MWLOW  BCUT2D_CHGHI  ...  fr_amide  fr_aniline  fr_bicyclic  \\\n",
              "0            -0.24          0.00  ...         0           0            0   \n",
              "1            -0.19         -0.00  ...         0           0            0   \n",
              "2            -0.24          0.00  ...         1           0            0   \n",
              "3            -0.14          0.51  ...         0           1            0   \n",
              "4            -0.25         -0.04  ...         0           0            0   \n",
              "...            ...           ...  ...       ...         ...          ...   \n",
              "2868         -0.18         -0.00  ...         0           0            0   \n",
              "2869         -0.09          0.55  ...         0           0            1   \n",
              "2870         -0.25         -0.00  ...         1           1            1   \n",
              "2871         -0.09          0.51  ...         0           0            1   \n",
              "2872         -0.25         -0.00  ...         0           0            0   \n",
              "\n",
              "      fr_ester  fr_ether  fr_halogen  fr_ketone  fr_ketone_Topliss  \\\n",
              "0            0         0           3          0                  0   \n",
              "1            0         0           1          0                  0   \n",
              "2            0         1           1          0                  0   \n",
              "3            0         0           2          0                  0   \n",
              "4            0         0           0          0                  0   \n",
              "...        ...       ...         ...        ...                ...   \n",
              "2868         0         0           0          0                  0   \n",
              "2869         0         0           1          0                  0   \n",
              "2870         0         2           0          0                  0   \n",
              "2871         0         0           0          0                  0   \n",
              "2872         0         0           0          0                  0   \n",
              "\n",
              "      fr_methoxy  fr_para_hydroxylation  \n",
              "0              0                      0  \n",
              "1              0                      0  \n",
              "2              0                      0  \n",
              "3              0                      0  \n",
              "4              0                      0  \n",
              "...          ...                    ...  \n",
              "2868           0                      0  \n",
              "2869           0                      0  \n",
              "2870           0                      1  \n",
              "2871           0                      0  \n",
              "2872           0                      1  \n",
              "\n",
              "[2873 rows x 106 columns]"
            ]
          },
          "execution_count": 87,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# adapted from:https://github.com/dataprofessor/padel/blob/main/padelpy.ipynb\n",
        "\n",
        "from sklearn.feature_selection import VarianceThreshold\n",
        "\n",
        "def remove_low_variance(input_data, threshold=0.1):\n",
        "    selection = VarianceThreshold(threshold)\n",
        "    selection.fit(input_data)\n",
        "    return input_data[input_data.columns[selection.get_support(indices=True)]]\n",
        "\n",
        "X = remove_low_variance(descriptors_new, threshold=0.1)\n",
        "X"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 88,
      "id": "QBxEwW-3dlvC",
      "metadata": {
        "id": "QBxEwW-3dlvC"
      },
      "outputs": [],
      "source": [
        "# Assign Features and Labels\n",
        "X  # molecular features\n",
        "y = dataset_new.Energygap # labels or targets"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "wEZH7_9CdlvC",
      "metadata": {
        "id": "wEZH7_9CdlvC"
      },
      "source": [
        "## 2.  Train-test data split"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 89,
      "id": "boyasHRcdlvC",
      "metadata": {
        "id": "boyasHRcdlvC"
      },
      "outputs": [],
      "source": [
        "X_train, X_test, y_train, y_test = train_test_split(X,y,test_size=0.3,random_state=42)"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "99nepkDodlvD",
      "metadata": {
        "id": "99nepkDodlvD"
      },
      "source": [
        "## 3. Training and Selection of the best ML models for this problem"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 90,
      "id": "TR7P1icfdlvD",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TR7P1icfdlvD",
        "outputId": "37d2e20a-c455-4047-b482-dc28ebb3e073"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  0%|          | 0/42 [00:00<?, ?it/s]"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 98%|█████████▊| 41/42 [04:28<00:05,  5.43s/it]"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004066 seconds.\n",
            "You can set `force_col_wise=true` to remove the overhead.\n",
            "[LightGBM] [Info] Total Bins 11561\n",
            "[LightGBM] [Info] Number of data points in the train set: 2011, number of used features: 106\n",
            "[LightGBM] [Info] Start training from score 190.710045\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "100%|██████████| 42/42 [04:29<00:00,  6.42s/it]\n"
          ]
        }
      ],
      "source": [
        "# Sklearn ML models in 2 lines of code\n",
        "# 42 regression ML models\n",
        "\n",
        "lregs = LazyRegressor(verbose=0,ignore_warnings=True, custom_metric=None)\n",
        "models, prediction_tests = lregs.fit(X_train, X_test, y_train, y_test)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 91,
      "id": "XYjCTm7hdlvD",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "XYjCTm7hdlvD",
        "outputId": "a3989b52-9840-41d7-88b3-2154e679579d"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Adjusted R-Squared</th>\n",
              "      <th>R-Squared</th>\n",
              "      <th>RMSE</th>\n",
              "      <th>Time Taken</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Model</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>LGBMRegressor</th>\n",
              "      <td>0.90</td>\n",
              "      <td>0.91</td>\n",
              "      <td>9.00</td>\n",
              "      <td>0.97</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>HistGradientBoostingRegressor</th>\n",
              "      <td>0.90</td>\n",
              "      <td>0.91</td>\n",
              "      <td>9.10</td>\n",
              "      <td>2.84</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>ExtraTreesRegressor</th>\n",
              "      <td>0.89</td>\n",
              "      <td>0.91</td>\n",
              "      <td>9.35</td>\n",
              "      <td>8.78</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>XGBRegressor</th>\n",
              "      <td>0.88</td>\n",
              "      <td>0.90</td>\n",
              "      <td>9.77</td>\n",
              "      <td>2.03</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>RandomForestRegressor</th>\n",
              "      <td>0.87</td>\n",
              "      <td>0.89</td>\n",
              "      <td>10.28</td>\n",
              "      <td>24.06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>GradientBoostingRegressor</th>\n",
              "      <td>0.87</td>\n",
              "      <td>0.88</td>\n",
              "      <td>10.43</td>\n",
              "      <td>7.09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>BaggingRegressor</th>\n",
              "      <td>0.85</td>\n",
              "      <td>0.87</td>\n",
              "      <td>10.91</td>\n",
              "      <td>2.57</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>PoissonRegressor</th>\n",
              "      <td>0.83</td>\n",
              "      <td>0.85</td>\n",
              "      <td>11.85</td>\n",
              "      <td>0.11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>LassoLarsIC</th>\n",
              "      <td>0.82</td>\n",
              "      <td>0.84</td>\n",
              "      <td>12.00</td>\n",
              "      <td>0.26</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>TransformedTargetRegressor</th>\n",
              "      <td>0.82</td>\n",
              "      <td>0.84</td>\n",
              "      <td>12.00</td>\n",
              "      <td>0.06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>LinearRegression</th>\n",
              "      <td>0.82</td>\n",
              "      <td>0.84</td>\n",
              "      <td>12.00</td>\n",
              "      <td>0.07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Ridge</th>\n",
              "      <td>0.82</td>\n",
              "      <td>0.84</td>\n",
              "      <td>12.01</td>\n",
              "      <td>0.06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>RidgeCV</th>\n",
              "      <td>0.82</td>\n",
              "      <td>0.84</td>\n",
              "      <td>12.01</td>\n",
              "      <td>0.08</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>LassoCV</th>\n",
              "      <td>0.82</td>\n",
              "      <td>0.84</td>\n",
              "      <td>12.10</td>\n",
              "      <td>1.62</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>LassoLarsCV</th>\n",
              "      <td>0.82</td>\n",
              "      <td>0.84</td>\n",
              "      <td>12.12</td>\n",
              "      <td>0.76</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>BayesianRidge</th>\n",
              "      <td>0.82</td>\n",
              "      <td>0.84</td>\n",
              "      <td>12.15</td>\n",
              "      <td>0.15</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>SGDRegressor</th>\n",
              "      <td>0.82</td>\n",
              "      <td>0.84</td>\n",
              "      <td>12.25</td>\n",
              "      <td>0.10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>ElasticNetCV</th>\n",
              "      <td>0.81</td>\n",
              "      <td>0.83</td>\n",
              "      <td>12.38</td>\n",
              "      <td>1.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>HuberRegressor</th>\n",
              "      <td>0.81</td>\n",
              "      <td>0.83</td>\n",
              "      <td>12.39</td>\n",
              "      <td>0.44</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>LinearSVR</th>\n",
              "      <td>0.80</td>\n",
              "      <td>0.82</td>\n",
              "      <td>12.84</td>\n",
              "      <td>0.30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Lasso</th>\n",
              "      <td>0.77</td>\n",
              "      <td>0.80</td>\n",
              "      <td>13.65</td>\n",
              "      <td>0.08</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>LassoLars</th>\n",
              "      <td>0.77</td>\n",
              "      <td>0.80</td>\n",
              "      <td>13.65</td>\n",
              "      <td>0.08</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>AdaBoostRegressor</th>\n",
              "      <td>0.77</td>\n",
              "      <td>0.80</td>\n",
              "      <td>13.73</td>\n",
              "      <td>2.45</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>OrthogonalMatchingPursuit</th>\n",
              "      <td>0.76</td>\n",
              "      <td>0.79</td>\n",
              "      <td>13.87</td>\n",
              "      <td>0.06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>ExtraTreeRegressor</th>\n",
              "      <td>0.76</td>\n",
              "      <td>0.79</td>\n",
              "      <td>14.03</td>\n",
              "      <td>0.12</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>DecisionTreeRegressor</th>\n",
              "      <td>0.75</td>\n",
              "      <td>0.78</td>\n",
              "      <td>14.11</td>\n",
              "      <td>0.31</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>OrthogonalMatchingPursuitCV</th>\n",
              "      <td>0.75</td>\n",
              "      <td>0.78</td>\n",
              "      <td>14.18</td>\n",
              "      <td>0.10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>ElasticNet</th>\n",
              "      <td>0.75</td>\n",
              "      <td>0.78</td>\n",
              "      <td>14.19</td>\n",
              "      <td>0.07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>GammaRegressor</th>\n",
              "      <td>0.75</td>\n",
              "      <td>0.78</td>\n",
              "      <td>14.25</td>\n",
              "      <td>0.07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>TweedieRegressor</th>\n",
              "      <td>0.74</td>\n",
              "      <td>0.77</td>\n",
              "      <td>14.48</td>\n",
              "      <td>0.08</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>KNeighborsRegressor</th>\n",
              "      <td>0.74</td>\n",
              "      <td>0.77</td>\n",
              "      <td>14.66</td>\n",
              "      <td>0.21</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>PassiveAggressiveRegressor</th>\n",
              "      <td>0.69</td>\n",
              "      <td>0.73</td>\n",
              "      <td>15.80</td>\n",
              "      <td>0.09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>LarsCV</th>\n",
              "      <td>0.67</td>\n",
              "      <td>0.71</td>\n",
              "      <td>16.28</td>\n",
              "      <td>0.47</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>SVR</th>\n",
              "      <td>0.64</td>\n",
              "      <td>0.69</td>\n",
              "      <td>17.03</td>\n",
              "      <td>1.68</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>NuSVR</th>\n",
              "      <td>0.62</td>\n",
              "      <td>0.66</td>\n",
              "      <td>17.63</td>\n",
              "      <td>1.07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>MLPRegressor</th>\n",
              "      <td>0.56</td>\n",
              "      <td>0.61</td>\n",
              "      <td>18.90</td>\n",
              "      <td>4.84</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>DummyRegressor</th>\n",
              "      <td>-0.14</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>30.43</td>\n",
              "      <td>0.06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>QuantileRegressor</th>\n",
              "      <td>-0.18</td>\n",
              "      <td>-0.04</td>\n",
              "      <td>30.99</td>\n",
              "      <td>200.70</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>GaussianProcessRegressor</th>\n",
              "      <td>-37.21</td>\n",
              "      <td>-32.50</td>\n",
              "      <td>176.13</td>\n",
              "      <td>2.50</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>KernelRidge</th>\n",
              "      <td>-44.00</td>\n",
              "      <td>-38.46</td>\n",
              "      <td>191.14</td>\n",
              "      <td>0.30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>RANSACRegressor</th>\n",
              "      <td>-92.61</td>\n",
              "      <td>-81.08</td>\n",
              "      <td>275.68</td>\n",
              "      <td>0.74</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Lars</th>\n",
              "      <td>-193697010274448896.00</td>\n",
              "      <td>-169850456164005696.00</td>\n",
              "      <td>12540561752.48</td>\n",
              "      <td>0.12</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                  Adjusted R-Squared              R-Squared  \\\n",
              "Model                                                                         \n",
              "LGBMRegressor                                   0.90                   0.91   \n",
              "HistGradientBoostingRegressor                   0.90                   0.91   \n",
              "ExtraTreesRegressor                             0.89                   0.91   \n",
              "XGBRegressor                                    0.88                   0.90   \n",
              "RandomForestRegressor                           0.87                   0.89   \n",
              "GradientBoostingRegressor                       0.87                   0.88   \n",
              "BaggingRegressor                                0.85                   0.87   \n",
              "PoissonRegressor                                0.83                   0.85   \n",
              "LassoLarsIC                                     0.82                   0.84   \n",
              "TransformedTargetRegressor                      0.82                   0.84   \n",
              "LinearRegression                                0.82                   0.84   \n",
              "Ridge                                           0.82                   0.84   \n",
              "RidgeCV                                         0.82                   0.84   \n",
              "LassoCV                                         0.82                   0.84   \n",
              "LassoLarsCV                                     0.82                   0.84   \n",
              "BayesianRidge                                   0.82                   0.84   \n",
              "SGDRegressor                                    0.82                   0.84   \n",
              "ElasticNetCV                                    0.81                   0.83   \n",
              "HuberRegressor                                  0.81                   0.83   \n",
              "LinearSVR                                       0.80                   0.82   \n",
              "Lasso                                           0.77                   0.80   \n",
              "LassoLars                                       0.77                   0.80   \n",
              "AdaBoostRegressor                               0.77                   0.80   \n",
              "OrthogonalMatchingPursuit                       0.76                   0.79   \n",
              "ExtraTreeRegressor                              0.76                   0.79   \n",
              "DecisionTreeRegressor                           0.75                   0.78   \n",
              "OrthogonalMatchingPursuitCV                     0.75                   0.78   \n",
              "ElasticNet                                      0.75                   0.78   \n",
              "GammaRegressor                                  0.75                   0.78   \n",
              "TweedieRegressor                                0.74                   0.77   \n",
              "KNeighborsRegressor                             0.74                   0.77   \n",
              "PassiveAggressiveRegressor                      0.69                   0.73   \n",
              "LarsCV                                          0.67                   0.71   \n",
              "SVR                                             0.64                   0.69   \n",
              "NuSVR                                           0.62                   0.66   \n",
              "MLPRegressor                                    0.56                   0.61   \n",
              "DummyRegressor                                 -0.14                  -0.00   \n",
              "QuantileRegressor                              -0.18                  -0.04   \n",
              "GaussianProcessRegressor                      -37.21                 -32.50   \n",
              "KernelRidge                                   -44.00                 -38.46   \n",
              "RANSACRegressor                               -92.61                 -81.08   \n",
              "Lars                          -193697010274448896.00 -169850456164005696.00   \n",
              "\n",
              "                                        RMSE  Time Taken  \n",
              "Model                                                     \n",
              "LGBMRegressor                           9.00        0.97  \n",
              "HistGradientBoostingRegressor           9.10        2.84  \n",
              "ExtraTreesRegressor                     9.35        8.78  \n",
              "XGBRegressor                            9.77        2.03  \n",
              "RandomForestRegressor                  10.28       24.06  \n",
              "GradientBoostingRegressor              10.43        7.09  \n",
              "BaggingRegressor                       10.91        2.57  \n",
              "PoissonRegressor                       11.85        0.11  \n",
              "LassoLarsIC                            12.00        0.26  \n",
              "TransformedTargetRegressor             12.00        0.06  \n",
              "LinearRegression                       12.00        0.07  \n",
              "Ridge                                  12.01        0.06  \n",
              "RidgeCV                                12.01        0.08  \n",
              "LassoCV                                12.10        1.62  \n",
              "LassoLarsCV                            12.12        0.76  \n",
              "BayesianRidge                          12.15        0.15  \n",
              "SGDRegressor                           12.25        0.10  \n",
              "ElasticNetCV                           12.38        1.00  \n",
              "HuberRegressor                         12.39        0.44  \n",
              "LinearSVR                              12.84        0.30  \n",
              "Lasso                                  13.65        0.08  \n",
              "LassoLars                              13.65        0.08  \n",
              "AdaBoostRegressor                      13.73        2.45  \n",
              "OrthogonalMatchingPursuit              13.87        0.06  \n",
              "ExtraTreeRegressor                     14.03        0.12  \n",
              "DecisionTreeRegressor                  14.11        0.31  \n",
              "OrthogonalMatchingPursuitCV            14.18        0.10  \n",
              "ElasticNet                             14.19        0.07  \n",
              "GammaRegressor                         14.25        0.07  \n",
              "TweedieRegressor                       14.48        0.08  \n",
              "KNeighborsRegressor                    14.66        0.21  \n",
              "PassiveAggressiveRegressor             15.80        0.09  \n",
              "LarsCV                                 16.28        0.47  \n",
              "SVR                                    17.03        1.68  \n",
              "NuSVR                                  17.63        1.07  \n",
              "MLPRegressor                           18.90        4.84  \n",
              "DummyRegressor                         30.43        0.06  \n",
              "QuantileRegressor                      30.99      200.70  \n",
              "GaussianProcessRegressor              176.13        2.50  \n",
              "KernelRidge                           191.14        0.30  \n",
              "RANSACRegressor                       275.68        0.74  \n",
              "Lars                          12540561752.48        0.12  "
            ]
          },
          "execution_count": 91,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "prediction_tests"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "Q9uzc5RZdlvD",
      "metadata": {
        "id": "Q9uzc5RZdlvD"
      },
      "source": [
        "# Parameter tuning of the top two models"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 92,
      "id": "yydCfnBkdlvD",
      "metadata": {
        "id": "yydCfnBkdlvD"
      },
      "outputs": [],
      "source": [
        "from sklearn.ensemble import HistGradientBoostingRegressor\n",
        "from lightgbm import LGBMRegressor\n",
        "\n",
        "lgbm = LGBMRegressor()\n",
        "gbr = HistGradientBoostingRegressor()"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "iMOVxjWHdlvE",
      "metadata": {
        "id": "iMOVxjWHdlvE"
      },
      "source": [
        "## Use the optimized parameters to predict test sets\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 93,
      "id": "b_ouefQFdlvE",
      "metadata": {
        "id": "b_ouefQFdlvE"
      },
      "outputs": [],
      "source": [
        "# Define LGBMRegressor model\n",
        "lgbm_opt =LGBMRegressor(max_depth=5,min_samples_split=4, n_estimators= 800,max_features='auto')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 94,
      "id": "3i8oqXP3dlvE",
      "metadata": {
        "id": "3i8oqXP3dlvE"
      },
      "outputs": [],
      "source": [
        "gbr_opt = HistGradientBoostingRegressor(learning_rate= 0.22)#learning_rate= 0.22"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 95,
      "id": "l2FDKarqdlvE",
      "metadata": {
        "id": "l2FDKarqdlvE"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "[LightGBM] [Warning] Unknown parameter: min_samples_split\n",
            "[LightGBM] [Warning] Unknown parameter: max_features\n",
            "[LightGBM] [Warning] Accuracy may be bad since you didn't explicitly set num_leaves OR 2^max_depth > num_leaves. (num_leaves=31).\n",
            "[LightGBM] [Warning] Unknown parameter: min_samples_split\n",
            "[LightGBM] [Warning] Unknown parameter: max_features\n",
            "[LightGBM] [Warning] Accuracy may be bad since you didn't explicitly set num_leaves OR 2^max_depth > num_leaves. (num_leaves=31).\n",
            "[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.007356 seconds.\n",
            "You can set `force_col_wise=true` to remove the overhead.\n",
            "[LightGBM] [Info] Total Bins 11512\n",
            "[LightGBM] [Info] Number of data points in the train set: 2011, number of used features: 106\n",
            "[LightGBM] [Info] Start training from score 190.710045\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Unknown parameter: min_samples_split\n",
            "[LightGBM] [Warning] Unknown parameter: max_features\n",
            "[LightGBM] [Warning] Accuracy may be bad since you didn't explicitly set num_leaves OR 2^max_depth > num_leaves. (num_leaves=31).\n"
          ]
        }
      ],
      "source": [
        "# Train the model and predict the energy gap using X_test \n",
        "lgbm_opt.fit(X_train, y_train)\n",
        "lgbm_predict = lgbm_opt.predict(X_test)\n",
        "\n",
        "# Train the model and predict the energy gap using X_test \n",
        "gbr_opt.fit(X_train, y_train)\n",
        "gbr_predict = gbr_opt.predict(X_test)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 96,
      "id": "vRlRbHzldlvF",
      "metadata": {
        "id": "vRlRbHzldlvF"
      },
      "outputs": [],
      "source": [
        "# Take the average of the optimized predictions\n",
        "ave_predict = (lgbm_predict + gbr_predict)/2"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 97,
      "id": "3QJsEbB1dlvF",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3QJsEbB1dlvF",
        "outputId": "48a84929-e4ff-4180-85da-a8f2d9e6018d"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "MAE: 6.031669675212837\n"
          ]
        }
      ],
      "source": [
        "# Model performance using MAE\n",
        "MAE = mean_absolute_error(y_test, ave_predict)\n",
        "print('MAE:', MAE)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 98,
      "id": "8_Sf-l6bdlvF",
      "metadata": {
        "id": "8_Sf-l6bdlvF"
      },
      "outputs": [],
      "source": [
        "df_ = pd.DataFrame()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 99,
      "id": "k2gp-o7ZdlvF",
      "metadata": {
        "id": "k2gp-o7ZdlvF"
      },
      "outputs": [],
      "source": [
        "df_['Actual'] = np.array(y_test)\n",
        "df_['Predicted'] = ave_predict"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 100,
      "id": "qodX-z3PdlvF",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "qodX-z3PdlvF",
        "outputId": "bd444066-8718-4afc-a93a-db0d2fb65d44"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Actual</th>\n",
              "      <th>Predicted</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>189.29</td>\n",
              "      <td>187.54</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>199.23</td>\n",
              "      <td>198.04</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>169.53</td>\n",
              "      <td>168.98</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>219.88</td>\n",
              "      <td>217.79</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>211.16</td>\n",
              "      <td>201.63</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>179.70</td>\n",
              "      <td>186.89</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>177.46</td>\n",
              "      <td>182.56</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>193.01</td>\n",
              "      <td>200.93</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>253.85</td>\n",
              "      <td>259.68</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>174.26</td>\n",
              "      <td>180.03</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>129.66</td>\n",
              "      <td>145.10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>157.71</td>\n",
              "      <td>153.62</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>167.70</td>\n",
              "      <td>168.18</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>166.17</td>\n",
              "      <td>169.97</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>196.27</td>\n",
              "      <td>199.20</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>169.38</td>\n",
              "      <td>168.08</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>159.52</td>\n",
              "      <td>165.24</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17</th>\n",
              "      <td>222.86</td>\n",
              "      <td>227.85</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18</th>\n",
              "      <td>162.81</td>\n",
              "      <td>161.73</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19</th>\n",
              "      <td>152.51</td>\n",
              "      <td>146.87</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>20</th>\n",
              "      <td>230.60</td>\n",
              "      <td>232.59</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>21</th>\n",
              "      <td>166.14</td>\n",
              "      <td>163.88</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>22</th>\n",
              "      <td>213.44</td>\n",
              "      <td>213.12</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>23</th>\n",
              "      <td>229.35</td>\n",
              "      <td>234.55</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24</th>\n",
              "      <td>166.33</td>\n",
              "      <td>171.78</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25</th>\n",
              "      <td>152.42</td>\n",
              "      <td>155.53</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>26</th>\n",
              "      <td>180.82</td>\n",
              "      <td>173.87</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>27</th>\n",
              "      <td>229.43</td>\n",
              "      <td>225.62</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>28</th>\n",
              "      <td>166.37</td>\n",
              "      <td>184.41</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>29</th>\n",
              "      <td>236.48</td>\n",
              "      <td>232.02</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>30</th>\n",
              "      <td>238.37</td>\n",
              "      <td>241.02</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>31</th>\n",
              "      <td>157.53</td>\n",
              "      <td>163.46</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32</th>\n",
              "      <td>233.65</td>\n",
              "      <td>251.60</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>33</th>\n",
              "      <td>168.75</td>\n",
              "      <td>161.42</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>34</th>\n",
              "      <td>157.11</td>\n",
              "      <td>170.22</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>35</th>\n",
              "      <td>167.12</td>\n",
              "      <td>166.84</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>36</th>\n",
              "      <td>175.38</td>\n",
              "      <td>197.90</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>37</th>\n",
              "      <td>215.71</td>\n",
              "      <td>218.75</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>38</th>\n",
              "      <td>186.96</td>\n",
              "      <td>185.91</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>39</th>\n",
              "      <td>220.83</td>\n",
              "      <td>222.37</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>40</th>\n",
              "      <td>211.70</td>\n",
              "      <td>218.23</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>41</th>\n",
              "      <td>185.26</td>\n",
              "      <td>193.22</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>42</th>\n",
              "      <td>184.95</td>\n",
              "      <td>184.35</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>43</th>\n",
              "      <td>192.10</td>\n",
              "      <td>194.85</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>44</th>\n",
              "      <td>249.58</td>\n",
              "      <td>242.33</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>45</th>\n",
              "      <td>160.26</td>\n",
              "      <td>156.40</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>46</th>\n",
              "      <td>126.77</td>\n",
              "      <td>151.42</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>47</th>\n",
              "      <td>175.03</td>\n",
              "      <td>176.24</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>48</th>\n",
              "      <td>170.68</td>\n",
              "      <td>173.03</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>49</th>\n",
              "      <td>177.83</td>\n",
              "      <td>176.88</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "    Actual  Predicted\n",
              "0   189.29     187.54\n",
              "1   199.23     198.04\n",
              "2   169.53     168.98\n",
              "3   219.88     217.79\n",
              "4   211.16     201.63\n",
              "5   179.70     186.89\n",
              "6   177.46     182.56\n",
              "7   193.01     200.93\n",
              "8   253.85     259.68\n",
              "9   174.26     180.03\n",
              "10  129.66     145.10\n",
              "11  157.71     153.62\n",
              "12  167.70     168.18\n",
              "13  166.17     169.97\n",
              "14  196.27     199.20\n",
              "15  169.38     168.08\n",
              "16  159.52     165.24\n",
              "17  222.86     227.85\n",
              "18  162.81     161.73\n",
              "19  152.51     146.87\n",
              "20  230.60     232.59\n",
              "21  166.14     163.88\n",
              "22  213.44     213.12\n",
              "23  229.35     234.55\n",
              "24  166.33     171.78\n",
              "25  152.42     155.53\n",
              "26  180.82     173.87\n",
              "27  229.43     225.62\n",
              "28  166.37     184.41\n",
              "29  236.48     232.02\n",
              "30  238.37     241.02\n",
              "31  157.53     163.46\n",
              "32  233.65     251.60\n",
              "33  168.75     161.42\n",
              "34  157.11     170.22\n",
              "35  167.12     166.84\n",
              "36  175.38     197.90\n",
              "37  215.71     218.75\n",
              "38  186.96     185.91\n",
              "39  220.83     222.37\n",
              "40  211.70     218.23\n",
              "41  185.26     193.22\n",
              "42  184.95     184.35\n",
              "43  192.10     194.85\n",
              "44  249.58     242.33\n",
              "45  160.26     156.40\n",
              "46  126.77     151.42\n",
              "47  175.03     176.24\n",
              "48  170.68     173.03\n",
              "49  177.83     176.88"
            ]
          },
          "execution_count": 100,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df_[:50]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 101,
      "id": "O3_Mj_d5dlvF",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "O3_Mj_d5dlvF",
        "outputId": "5c65a9b6-0ec3-4973-95b8-8310e1fd02f9"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "R^2: 0.9215413724552036\n"
          ]
        }
      ],
      "source": [
        "# R^2 (coefficient of determination) regression score function: \n",
        "R2 =r2_score(y_test, ave_predict)\n",
        "print('R^2:', R2)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 104,
      "id": "H5OeRFY3dlvF",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 313
        },
        "id": "H5OeRFY3dlvF",
        "outputId": "6ba231f6-305d-4865-888a-bedbdfda7a70",
        "scrolled": true
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "sn.regplot(x=ave_predict , y=y_test,line_kws={\"lw\":2,'ls':'--','color':'black',\"alpha\":0.7})\n",
        "plt.xlabel('Predicted Energy gap', color='blue')\n",
        "plt.ylabel('Observed Energy gap', color ='blue')\n",
        "plt.title(\"Test set\", color='red')\n",
        "plt.grid(alpha=0.2)\n",
        "\n",
        "# Format R2 and MAE as strings\n",
        "R2_str = \"{:04.2f}\".format(R2)\n",
        "MAE_str = \"{:04.2f}\".format(MAE)\n",
        "\n",
        "# Create legend labels\n",
        "R2_label = \"R2={}\".format(R2_str)\n",
        "MAE_label = \"MAE={}\".format(MAE_str)\n",
        "\n",
        "# Add legend\n",
        "plt.legend(handles=[mpatches.Patch(label=R2_label), mpatches.Patch(label=MAE_label)])\n",
        "\n",
        "# Save the plot as an SVG image\n",
        "plt.savefig('plot.svg', format='svg')\n",
        "\n",
        "# Show the plot\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 60,
      "id": "877588f8",
      "metadata": {},
      "outputs": [
        {
          "ename": "TypeError",
          "evalue": "unsupported format string passed to Patch.__format__",
          "output_type": "error",
          "traceback": [
            "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[1;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
            "Cell \u001b[1;32mIn[60], line 6\u001b[0m\n\u001b[0;32m      4\u001b[0m plt\u001b[39m.\u001b[39mtitle(\u001b[39m\"\u001b[39m\u001b[39mTest set\u001b[39m\u001b[39m\"\u001b[39m, color\u001b[39m=\u001b[39m\u001b[39m'\u001b[39m\u001b[39mred\u001b[39m\u001b[39m'\u001b[39m)\n\u001b[0;32m      5\u001b[0m plt\u001b[39m.\u001b[39mgrid(alpha\u001b[39m=\u001b[39m\u001b[39m0.2\u001b[39m)\n\u001b[1;32m----> 6\u001b[0m R2 \u001b[39m=\u001b[39m mpatches\u001b[39m.\u001b[39mPatch(label\u001b[39m=\u001b[39m\u001b[39m\"\u001b[39m\u001b[39mR2=\u001b[39m\u001b[39m{:04.2f}\u001b[39;00m\u001b[39m\"\u001b[39m\u001b[39m.\u001b[39mformat(R2))\n\u001b[0;32m      7\u001b[0m MAE \u001b[39m=\u001b[39m mpatches\u001b[39m.\u001b[39mPatch(label\u001b[39m=\u001b[39m\u001b[39m\"\u001b[39m\u001b[39mMAE=\u001b[39m\u001b[39m{:04.2f}\u001b[39;00m\u001b[39m\"\u001b[39m\u001b[39m.\u001b[39mformat(MAE))\n\u001b[0;32m      8\u001b[0m plt\u001b[39m.\u001b[39mlegend(handles\u001b[39m=\u001b[39m[R2, MAE])\n",
            "\u001b[1;31mTypeError\u001b[0m: unsupported format string passed to Patch.__format__"
          ]
        },
        {
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjsAAAHFCAYAAAAUpjivAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAADb50lEQVR4nOy9eZxcVZn//z53qb2q9yU72YCEhH2NQFA2ARFwYfzCOOKgg4ioA44K4g+dUXBHB0d0RkRmFHF0ZHFhRxIhIBAEkhDICtm603tV136X8/vjVlV3dVcnXZ3upOmc9+vV0F11695zqjt9P/0sn0dIKSUKhUKhUCgUUxTtQC9AoVAoFAqFYiJRYkehUCgUCsWURokdhUKhUCgUUxoldhQKhUKhUExplNhRKBQKhUIxpVFiR6FQKBQKxZRGiR2FQqFQKBRTGiV2FAqFQqFQTGmU2FEoFAqFQjGlUWJHoVBMPEKM7uOpp/b9Wuk0fOUr43Ouatm1y7v2yy/v/2srFIoRMQ70AhQKxUHAs8+Wf/1v/wZ//jM8+WT544sX7/u10mn46le9z884Y9/PVw27dnnXPuQQOPro/XtthUIxIkrsKBSKiefkk8u/bmoCTRv+uEKhUEwAKo2lUCgmB/k8fO1rcPjh4Pd7guijH4XOzvLjnnzSi9g0NEAwCLNnw/vf70V03nzTex14EZZieuyKK0a+rut61z3sMO98tbVw5JHwgx+UH7dxI1x2GTQ3e+tbtAj+4z8Gnn/qKTjhBO/zj3504Npf+cq+vCsKhWIcUJEdhUJx4HFduOgi+Mtf4POfh2XL4K234OabPWHz4oueEHnzTbjgAjjtNPjZzzxhsnMnPPywJ5amTfM+f/e74cor4WMf885fFECV+Na3PEFy001w+ulgWfD669DXN3DMa695a5o9G777XWhthUcegU9/Grq6vHUeeyzcdZcndG66yVsnwMyZE/KWKRSK0aPEjkKhOPD87/96IuX//g/e976Bx486youW/PzncPXVsHo1ZLPw7W97zxW57LKBz487zvv/zJmjS5M98wwsXVoegTn33PJjrrsOolF4+mmIxbzHzj4bcjn4xjc80VNXB0uWeM/Nn69SdArFJEKlsRQKxYHnD3/wojQXXgi2PfBx9NFeFKXYWXX00eDzwT/9E9x9N2zZsu/XPvFEeOUV+OQnvWhNIlH+fDYLTzwBl1wCoVD5+s4/33v+uef2fR0KhWLCUGJHoVAceHbv9tJGPh+YZvlHe7uXKgIvYvL4417dzDXXeF/Pnz+8vqYabrgBvvMdT7Ccd55XC3TmmV7qDKC72xM2t98+fG3nn+8dU1yfQqGYlKg0lkKhOPA0Nnoi4+GHKz8fjQ58ftpp3ofjeILk9tvhs5+Flhb40Ieqv7ZheGmq667zBNfjj8ONN3qprO3bvfSUrsOHP+wJrErMnVv9dRUKxX5DiR2FQnHgec974N57PQFz0kmje42ue8cefjj88pfw0kue2PH7veczmerXUVsLH/iAV/T82c96BdGLF8M73wl/+5vXpeXzjfz6fbm2QqGYMJTYUSgUB54PfcgTLOefD5/5jFdHY5qwY4dnPnjRRV7NzI9/7LWeX3CB1xmVzXpdWQBnneX9PxqFOXPggQe8dFR9vRc5OuSQyte+8EKvsPj4472urbfegu9/3zvHwoXeMT/4AZx6qhdRuvpq71z9/bBpE/z+9wPmiPPne11jv/yl15oeicD06d6HQqE4YKiaHYVCceDRdXjwQS999LvfecLm4ou9TqdAwOuWAq9A2ba9Vu/zzvNSS52d3mvPOWfgfHfe6RUTv/e9XjfXnrxu3vlOWLkSPvEJr8Pqpps8kbRihSe4wIvuvPSSJ4puusm71pVXwm9/6x1bJBTyxFd3t3fMCSfAf/7nOL9ZCoWiWoSUUh7oRSgUCoVCoVBMFCqyo1AoFAqFYkqjxI5CoVAoFIopjRI7CoVCoVAopjRK7CgUCoVCoZjSKLGjUCgUCoViSqPEjkKhUCgUiimNMhUEXNdl165dRKNRhBAHejkKhUKhUChGgZSS/v5+pk+fjqaNHL9RYgfYtWsXs2bNOtDLUCgUCoVCMQa2b9/OzJkzR3xeiR0gWhgyuH37dmKx2LifX0pJPB6npqZmykaO1B6nDgfDPtUepwZqj1ODfdljIpFg1qxZpfv4SCixA6U3NxaLTZjYkVISi8Wm9A+r2uPU4GDYp9rj1EDtcWowHnvc2+tUgbJCoVAoFIopzQEVO3fccQdHHnlkKaJyyimn8NBDD5Wel1Lyla98henTpxMMBjnjjDNYt25d2TlyuRzXXnstjY2NhMNh3vve97Jjx479vRWFQqFQKBSTlAMqdmbOnMk3vvENXnzxRV588UXe9a53cdFFF5UEzbe+9S2+973v8cMf/pAXXniB1tZWzj77bPr7+0vn+OxnP8t9993Hvffey9NPP00ymeQ973kPjuMcqG0pFAqFQqGYRBxQsXPhhRdy/vnnc+ihh3LooYfy9a9/nUgkwnPPPYeUku9///t86Utf4n3vex9Llizh7rvvJp1Oc8899wAQj8e58847+e53v8tZZ53FMcccwy9+8QvWrFnD448/fiC3plAoFAqFYpIwaQqUHcfhN7/5DalUilNOOYWtW7fS3t7OOeecUzrG7/ezfPlyVq1axVVXXcXq1auxLKvsmOnTp7NkyRJWrVrFueeeW/FauVyOXC5X+jqRSAADRVLjTfG8E3HuyYLa49ThYNin2uPUQO1xarAvexztaw642FmzZg2nnHIK2WyWSCTCfffdx+LFi1m1ahUALS0tZce3tLTw1ltvAdDe3o7P56Ourm7YMe3t7SNe89Zbb+WrX/3qsMfj8fiEiZ1kMgnsvWL87Yra49ThYNin2uPUQO1xarAveywGK/bGARc7hx12GC+//DJ9fX383//9Hx/5yEdYsWJF6fmhG5dS7vXN2NsxN9xwA9ddd13p62Kffk1NzYS1ngNT3icB1B6nAgfDPtUepwZqj1ODfdnjaI8/4GLH5/OxYMECAI4//nheeOEFfvCDH/CFL3wB8KI306ZNKx3f0dFRiva0traSz+fp7e0ti+50dHSwbNmyEa/p9/vx+/3DHhdCTNgPU/HcU/WHFdQepxIHwz7VHqcGao9Tg7HucbTHTzqfHSkluVyOuXPn0traymOPPVZ6Lp/Ps2LFipKQOe644zBNs+yYtrY21q5du0exo1AoFAqF4uDhgEZ2brzxRs477zxmzZpFf38/9957L0899RQPP/wwQgg++9nPcsstt7Bw4UIWLlzILbfcQigU4rLLLgO8kNeVV17J9ddfT0NDA/X19Xzuc59j6dKlnHXWWQdyawqFQqFQDMN1Jet2JehJ56kP+ThiegxNm7oRm8nCARU7u3fv5sMf/jBtbW3U1NRw5JFH8vDDD3P22WcD8PnPf55MJsMnP/lJent7Oemkk3j00UfLZmDcdtttGIbBpZdeSiaT4cwzz+TnP/85uq4fqG0pFAqFQjGMVZu6uGPFZjZ3JLEciakL5jdHuHr5fJYtaDzQy5vSCDmV+9lGSSKRoKamhng8rgaBjhG1x6nDwbBPtcepwdtpj6s2dXHjfWtI5mzqQj58ukbecelNW0T8OrdcsrSi4Hk77XGs7Osg0NHcvyddzY5CoVAoFFMJ15XcsWIzyZxNayxAwNTRNEHA1GmN+UnmHO5YsRnXPehjDxOGEjsKhUKhUEwg63Yl2NyRpC7kGxa5EEJQGzLZ3JFk3a7RecYoqkeJHYVCoVAoJpCedB7Lkfj0yrdcv65huZKedH4/r+zgQYkdhUKhUCgmkPqQD1MX5B234vM5x8XUBPUh335e2cGDEjsKhUKhUEwgR0yPMb85Qm/aGjaSSEpJX9pifnOEI6aPf4OMwkOJHYVCoVAoJhBNE1y9fD4Rv057IkfGcnBdScZyaE/kiPh1rl4+X/ntTCBK7CgUCoVCMcEsW9DILZcsZdG0KOmcTUcyRzpns2hadMS2c8X4ccBnYykUCoVCcTCwbEEjJ89rUA7KBwAldhQKhUKh2E9ommDpzJoDvYyDDpXGUigUCoVCMaVRYkehUCgUCsWURokdhUKhUCgUUxoldhQKhUKhUExplNhRKBQKhUIxpVFiR6FQKBQKxZRGiR2FQqFQKBRTGuWzo1AoFArFQYLryoPS1FCJHYVCoVAoDgJWberijhWb2dyRxHIkpi6Y3xzh6uXzp/y4CpXGUigUCoViirNqUxc33reG9W0Jwn6D5qifsN9gfVs/N963hlWbug70EicUJXYUCoVCoZjCuK7kjhWbSeZsWmMBAqaOpgkCpk5rzE8y53DHis24rjzQS50wlNhRKBQKhWIKs25Xgs0dSepCPoQor88RQlAbMtnckWTdrsQBWuHEo8SOQqFQKBRTmJ50HsuR+PTKt3y/rmG5kp50fj+vbP+hxI5CoVAoFFOY+pAPUxfkHbfi8znHxdQE9SHffl7Z/kOJHYVCoVAopjBHTI8xvzlCb9pCyvK6HCklfWmL+c0RjpgeO0ArnHiU2FEoFAqFYgqjaYKrl88n4tdpT+TIWA6uK8lYDu2JHBG/ztXL509pvx0ldhQKhUKhmOIsW9DILZcsZdG0KOmcTUcyRzpns2halFsuWTqhPjuWZZHL5Sbs/KNBmQoqFAqFQnEQsGxBIyfPa9hvDspSSvr6+ujq6qKlpQW/3z8h1xkNSuwoFAqFQnGQoGmCpTNrJvw62WyW3bt3H/CIThEldhQKhUKhUIwLruvS1dVFX1/fgV5KGUrsKBQKhUKh2GeSySQdHR3Yto2UcpiB4YFEiR2FQqFQKCYhritZuzNOZ08vTfWwZEbNpOyYsm2bjo4OkskkqVSK+++/H4DLL7/8wC5sEErsKBQKhUIxyShOKN/S0U+96dBj6cxrjk66CeXFAmTHcfjrX//Kb3/7W5LJJAAnn3wy8+fPP8Ar9Digree33norJ5xwAtFolObmZi6++GLeeOONsmOEEBU/vv3tb5eOOeOMM4Y9/6EPfWh/b0ehUCgUin1m6ITy+rBv0k0oz+fzbN++nY6ODnbt2sVtt93Gz3/+85LQ8fl8dHd3A6DrOj7fgXVnPqCRnRUrVnDNNddwwgknYNs2X/rSlzjnnHN47bXXCIfDALS1tZW95qGHHuLKK6/k/e9/f9njH//4x/nXf/3X0tfBYHDiN6BQKBSKtyWuK/dbC3a16xo8oVwTIIRTmFDumQLesWIzJ89rOCDrlVLS3d1Nb28vuVyOP/3pTzz22GM4jlM65phjjuGDH/wg9fX1xGIxmpqa0HV9v691MAdU7Dz88MNlX9911100NzezevVqTj/9dABaW1vLjnnggQd45zvfybx588oeD4VCw45VKBQKhWIoxRTR5o4kliMxdcH85sikSBENn1A+MN5h6ITy/dFCPph0Ok1HRwf5fJ61a9dy77330tU1EGVqaGjgQx/6EEuXLsXv99PS0kIgENivaxyJSVWzE4/HAaivr6/4/O7du/njH//I3XffPey5X/7yl/ziF7+gpaWF8847j5tvvploNFrxPLlcrqz3P5HwxtpLKYfNDRkPiuediHNPFtQepw4Hwz7VHqcGY9njqk1d3HT/GpI5m7qQD5+ukXdcXm9L8KX7XuVrF0+sm/De6E7lsB0Xvy4QSAQSSv+HgC5IuC7dqdx++946jkNnZ2fpXrl+/Xpuv/320vOGYXD22Wdz3nnnEQgEaGxspLa2FmBUa9yXn9XRvmbSiB0pJddddx2nnnoqS5YsqXjM3XffTTQa5X3ve1/Z45dffjlz586ltbWVtWvXcsMNN/DKK6/w2GOPVTzPrbfeyle/+tVhj8fj8QkTO8U85mRqxRtP1B6nDgfDPtUepwbV7tF1Jfc+8wZhYTGnsVhD4k0Cnx4y6ErmufeZN1jUYBywlFaEPDPCkqBu4TN0BJIaw8WL8QhytoMZkkTIlwIEE0kqlaK3t7csTXXIIYcwZ84ctm7dyoIFC7j00ktpbW0lFApRW1uLEKKqte3Lz2pRgO2NSSN2PvWpT/Hqq6/y9NNPj3jMz372My6//PJhYbGPf/zjpc+XLFnCwoULOf7443nppZc49thjh53nhhtu4Lrrrit9nUgkmDVrFjU1NcRi4z/1tSigampqpvQvHVB7nAocDPtUe5waVLvHNTvivNSWJez302MNryHJaYKX2rLsSO0fl+FKHBONEYnt4PW2flpjJprwElm9lo4roT1hcfi0GMcsmD6hgiyfz9PR0UE6naavr4+mpqay5z/84Q+zY8cOTjrpJAKBAM3NzWOuld2Xn9XRHj8pxM61117Lgw8+yMqVK5k5c2bFY/7yl7/wxhtv8Otf/3qv5zv22GMxTZONGzdWFDt+v7/ijI5iJ9dEMLhTbKqi9jh1OBj2qfY4Nahmj70Zi7wDtbqOZPjxPl2nz7XpzVgH7D3TdcHVyxdw431raEvkqQsZ1BqQtlx60zYRv8HVyxeg6xPTTC2lpLe3l+7ubpLJJPfffz9/+ctfuO666zj00ENLx82aNYs5c+aUpaz2hbH+rI72+APaei6l5FOf+hS/+93vePLJJ5k7d+6Ix955550cd9xxHHXUUXs977p167Asi2nTpo3nchUKhULxNqY+5MPUBXnHrfh8znExNUF96MC2SQ+dUN6Tzu+XCeWZTIa33nqLzs5Onn32WW6++Wb+8pe/AHDPPfdg23bp2Fgsxty5c8dF6OwPDmhk55prruGee+7hgQceIBqN0t7eDnihrMHhsEQiwW9+8xu++93vDjvH5s2b+eUvf8n5559PY2Mjr732Gtdffz3HHHMM73jHO/bbXhQKhUJRPfuzBfyI6THmN0dY39ZPa0wriwpIKelLWyyaFuWI6eNfzlAtxQnlAw7KdRPmoOy6Lp2dncTjcdrb2/nVr35V5nnn8/k49dRTEULg9/v3KWV1oDigYueOO+4APFPAwdx1111cccUVpa/vvfdepJT8v//3/4adw+fz8cQTT/CDH/yAZDLJrFmzuOCCC7j55psPeF+/QqFQKEZmf7eAa5rg6uXzufG+NbQnctSGTPy6Rs5x6UtbRPw6Vy+fPyn8dmBgQnk8OnG1V/39/XR0dJDJZHjooYd49NFHK3rmNDY20tDQQF1d3bivYX8g5FTuSxwliUSCmpoa4vH4hBUox+PxKV8oqPY4NTgY9qn2eOApugQPbQHvLYiO0aRrxrrHMpHlSkxt8vjsDGWivo+WZdHR0UEqlWLdunX86le/GtEzJxqN0tTUhGFMTHxkX/Y42vv3pChQVigUCsXBw1CX4OINLqDptMa0CXcJLqaIJqOD8kQzuAC56G3z8MMPl4SOruucffbZnH/++UQiEZqbmwmFQgd41fuOEjsKhUKh2K8MdwkeYH+5BBdTRAcT2WyW3bt3l5nqCiG47LLL+NrXvsa8efO47LLLmDFjBvX19dTV1U3KqOBYUGJHoVAoFPuVnnQey5H4Rmif9usacVfSk87v55VNTVzXpauri76+PrZu3YrrumXTyKdNm8YXv/hFZs6cWRrMPVEpqwPF1NqNQqFQKCY9g1vAA1oFc79J0gI+Fejv7y+Nerj//vtZuXIlzc3N3HTTTWWTyOfPn09TU1NpCPdUQ4kdhUKhUOxX3k4t4BPFRLfcFwuQk8kkL7zwAr/5zW/o7+8HoKOjg2eeeYZ3vvOdCCFKXVZTJWVVCSV2FAqFQrFfebu1gI83E9lyL6Wkr6+P7u5u2trauOeee4Z55lx44YWcdtppRCIRmpqaME1zX7c06VFiR6FQKBT7naJLcPGmHy+0gC+aFt0vLeD708xwMCO13K9v6+fG+9bsk0NysQC5v7+fhx9+mEceeaSiZ05LSwvNzc1TNmVVCSV2FAqFQnFAOFAt4PvbzLDIRLXcDy5A3rJlC3fddRednZ2l5+vr6/nQhz7EUUcdRX19PfX19VM6ZVUJJXYUCoVCccDY3y3g4xlZqTY6NBEt98UC5OLcqnA4TE9PDwCapnHOOedw/vnnU1dXR3Nz80GRsqqEEjsKhUKhOCgYz8jKWKJD49lyP9gBeTAtLS28+93vZsOGDVx22WXMnj37oEtZVUKJHYVCoVAcFIxXZGVP0aEb7lvDx0+bx6z60LBoz3i03BcLkLu6uti6dSuPPPIIH/3oR8vayM877zwuvPBCGhoaDsqUVSWU2FEoFArFQcF4RFb2FB2K+F129mX419+/Rixg4DO0smjPvrbcFwuQe3t7S545AK2trVx00UWl42pqag7qlFUlKn/HFQqFQqF4G+G6kjU74qzY0MmaHXFcd/iM68GRlUqMJrIyUnQombPZ1ZfFdSWulMSCJmG/UaoFWrWpq9RyH/HrtCdyZCwH15VkLIf2RG7ElnvXdeno6OCtt95i5cqV3HzzzSWhA7Bu3Tocx8EwDKZPn86MGTOU0BmCiuwoFAqF4m3Nqk1d/Hjllor1M4O7vWqDJvOawrzenhyzmWGl6JBE0tmfw5USQxc4LrhSEjaNYbVA1bbcJ5NJ2tra6O7u5le/+lVFz5x3vetdNDU10dDQoFJWI6DEjkKhUCjetryyvY+vPLyF/pwzrH7mn//3ZZqjfrqT+ZIIaoj40DXGbGZYqe4mm3fJ2Q564XVCSAxNK3w+vBZoNC33tm3T0dFBd3c3v//97/nzn/9c5plz9NFHc+mllzJz5kyam5vLanYUw1FiR6FQKBRvS1xX8tvVOwr1M8Fh9TPbetL0pPLMqQ/hN3TyjktbPIeuwbQaTwRVa2ZYqe7Gdl2kBITEdiFoagTMgchPpVqgPbXcFwuQbdvmm9/8Jtu3b0fTvGsVPXOOPfZYmpqaiEaj+/5GHgQosaNQKBSKtyXrdiXY0ZseVj8jkXQl8ww8ItA0UdZiXhP08bWLl9KXsaoyM6w06kITApDYDuiaRlM0ULae0Q42zefz7N69m0wmU7iWxvHHH8/27dvRdZ2zzz6bCy64gJaWFhoaGtA0VXY7WpTYUSgUCsV+YzzHNPSk89iOxGeW3/QHp5VcCbbrAl7KqZhW2tKZRBOC5Yc2VX3doXU3liPRNQ1XSqbXBoj4B26to6kFklLS3d1Nd3c3tm2XFRefddZZ7Ny5k/PPP5+FCxeqlNUYUWJHoVAoFPuF8R7TUB/yYeiClOPiHxTlKKaVhAYCSvUzRaox7xuJoXU323vS/NfKzSRzDoaujboWKJ1Os3v3bjZu3Mg999zDokWLuOSSS0rPm6bJFVdcwaxZs4jFpu4U+IlGiR2FQqFQTDjjPQBzcIv3mx05ptdqaMITNZ64kTgOBH16Wf0MQM5xQMLWrtQ+RZeG1t3MawyPusvKcRw6Oztpb2/ngQceYMWKFQDs2LGDE088kRkzZnhRqNpaDMN429bmHKiBq0NRYkehUCgUE8p4D8AsRoi2dPQT1Wz6sw4bdidpjgaoDZogJEIIXClpjPjL6mf6sxY7+zIIBD98YuMw4799uTmPdrBpIpGgo6ODv/71r/zv//4v/f39pedaW1txHIdgMEhLSwumaRKPx0d1/cnGgRq4WgkldhQKhUIxoYznAMzBEaL6kElLUCcpHXb352lPZEnmbMI+ncNbo3T058rSSn0Zi7a4V/w7rcZPbdBXFl26/KTZrNzYVbo5AzTH/Cxf2MQJ8+ppDPv3Kn721GWVz+fp6Ohg69at3HPPPRU9c84++2xaW1tLKSsph5sjvh0Y70jevqLEjkKhUCgmlPEagDk0QqQJEMKhPuynJuhjZ1+WmXVBvn7JUpbOqOG5Ld0DaSVHksha6JpgRm2QaMArAi5Gl7b1ZPjuYxuI+A38hk46b5GxHDqTOdbtSqD/ZTO1QZPF02uqjkxIKenp6aG9vZ2HHnqIRx55BMdxkFKSyjnMOXQxF7//g5x14lKamhrHpcvqQKaPxjuSNx4osaNQKBSKCWU8BmBCpQjRQNRD0zQao356Unk04bWaD04rvfRWL7f/eSO1QZOAqZPO26Tznklf0NCwHIe87WKGBJ39ORzXZfDECceFeMbile3xUmTi5HkNrNkZ5+VtfUgBx8yqZemMmrIbeCaTYffu3eTzeVatWsWf/vQnAOLpPJ2WSeCwM8g1Hc7tL+d5uGMrVy/X9jnicaDTR+MZyRsvlNhRKBQKxYSyrwMwi4wlQlRMK/Wk8wgEliPZ0ZckZ7kMTRAJoDdt4UpZek4IKGaSXAmO69Kftbn1ofVICRs7kliFWVuGpnFYa4QbzlvESXPr6OzsJJFIlM7/jne8g6eeeooNW3fQW7+Y6IITaG6dQSgcHZbiKYq07lSOCHmOicbQ9dHVMx3o9NF4RfLGEyV2FAqFQjGhVDLiq3ZMA+xbhKg+5MOVLjv68jgVhoSCFyeybImmgXRh6Gp04V076NNZt8sTMV5ruwABjuOybleCT//3M3zutFZiVjcLFiwovd4wDD7ykSv49qMbyRuNzGxtQdO9fQxO8dz60HpqgiZbOlPYjsuMsCQS28HVyxfsUahMlvTReEXyxhNlv6hQKBSKCadoxLdoWpR0zqYjmSOds1k0LTrqaEMxQtSbtoYV7hYjRPObIxUjRItaozguJaEjxMDHYFwpKSXIBj0nGIjy9GdtXOl9bhoamhAIBMJ1sBNd7NzwKjfe/K985zvfYfv27eUXqJlGf81cmpsHhE7pGkLgNzRea+tn7c44Yb9Bc9RP0NR5fdD09JGoJn00kezL92miUJEdhUKhUOwXRtuaPRJDI0R1IQNpSDKWQ2/a3mOEaH17f3naaoQmJ4mXrhKFLwans4pYdiFtpQtcFyzHwckksRKdZDY+R3b7Wvo1aJxWwz333MPnP/95DMOgsbGRNvI4or1iikci6UvncaWkJugjYOoIJD5DpzVm0pbI7zEyM1nSR+MVyRtPlNhRKBQKxX5jT63ZlRjcVVQbNAn7Df7uhFk8sm43HfE0PdIlndf3OsizeIPXAARlxcdCeI87xdqcwpMloYOXwnIkmJogO+jFuVwWO9lLbudrpF9/BjefAryCZsJ1fOADH6Curo6GhgZ0Xae+Pz5iiqc05kIIzCGCZTSFvZMpfTR0pEa1A1fHGyV2FAqFQjEpGdxVlMo7ZC0HKSFo6oT8OrGgj+NnBTly7nQuPGo6hjFyZUZ9oVhX0zzhIF1ZKjxGgosneIRWKEgeJIY0DVwkmhDomkAT4LouuUSCfM9OUutXYHd76SoBoJuE5p9A7B3vYvnyMwiFgqVz7alY23IcHBdCPm2Y6zPsPTIzXoXg48W+RvLGE1Wzo1AoFIpJR7GraH1bAiEE6ZyN7bg4rksqb9OdzLOpo58nXu/g639ax0fvfmGP9SxHTI9xWGvUK0J2PKEjoFSfI/EEz+EtEb763iM4YnqMmoCB4U2eQBOCWMDgmNm1HFpv4Pa10b/uz8RX/aokdADMprnUnvr3NBx9Fml/A5t7yoVJMcUT8eu0J3JkLAfX9VJx8ayNpnkRnKE1N7D3yMyezt2eyB2Q9FExkrf80CaWzqw5IEIHVGRHoVAoFJOMYldRf9aiJmiyK57FldKrkZEU3I0lvsKgz0ze5bVdiT22Vmua4BPL5/H81m4cytNURQSeX8/lJ83h8pPmeK3fyRy9aYvasEmtX6dRz/DM64KvvLWdno6tSNfxXheIElp0OoFZR+AL1zCtIUI671SMwoyU4lkyvYZ4Jk9bPIeUckyRmcmWPposHFCxc+utt/K73/2O119/nWAwyLJly/jmN7/JYYcdVjrmiiuu4O677y573UknncRzzz1X+jqXy/G5z32OX/3qV2QyGc4880x+9KMfMXPmzP22F4VCoVCMD+t2JXhtV5xM3iWeyWAXa2icQaknPGGiaxLLdWkNmsSzNnes2MyJh9Szvr2/VOcD0Jex6EnmiQYMkjnPQHBwTY7f1GgI++lO5ks1MYPrYnp7e+nu7iaTdzl2Th1Xnj6f27rOoPe53xE45ChCC5dh1DQRCgZorQmiawJTc+lJ5lmxoXNYCmekFM9zW7rLCnsDuiBnO7QnLCJ+Y1SRmcmUPposHFCxs2LFCq655hpOOOEEbNvmS1/6Eueccw6vvfYa4XC4dNy73/1u7rrrrtLXPl95CO+zn/0sv//977n33ntpaGjg+uuv5z3veQ+rV69G14cXaSkUCoVicuK6kt+u3k5v2iq0ew/coIfa4xRTUVKCIyW1IZPXdiX4wE+epSORJZVzyFgOQkDA1NEF9OdsptUEMHWt5KAcMnWCfh3pQkcyVxaNyeVytLW18fDDjyBiLYQbWqkJmlywZBp/2XA4r8eupKZlJkYoSsinEzR1MnmHtr4MmhB86+H12C4VXYwrFWsPjcwkXBczJDl8WmyvPjuDqbYQfKpzQMXOww8/XPb1XXfdRXNzM6tXr+b0008vPe73+2ltba14jng8zp133sn//M//cNZZZwHwi1/8glmzZvH4449z7rnnTtwGFAqFQjFurNrUxY+e2sRft/aUhI0YqUcccKSLKwVCeO7FluPSl85jOS7RgEHGsku+OmlXUhvy4biSXX1ZZtWHaIz4y86XdZxSTUw+b3PPX17jxdWree6xB9m+fRsy3EjzqX+HaejMrg9x+hGziJt1pCwI+w0sx2VzZ8orpMYTYrYraY758enaqF2MB0dmSg7KC6ajj9BSrtg7k6pmpzjGvr6+vuzxp556iubmZmpra1m+fDlf//rXaW5uBmD16tVYlsU555xTOn769OksWbKEVatWVRQ7uVyOXC5X+rpo5y2lnJAJs8Xzvl2n144Gtcepw8GwT7XHyceqTV3cdP8a+tJ5kBKNwsgGOdzJuIjrSmxXYmgaARN2dWYBSVPEpKMw38rQPLnkui7pXJ6Q6UV02vpSzKwLEfTpCARSSuLpPIdPi/LUmjf54B+fp+OVFWS2raFY3WPkdhNItRNsnc+OfJA/bkrz9ycfwoqNXazdGac/ayElFLNFhiawHYf2vgwz6oJMi/kKLsabOGlu/R7TSkLAkhkxb13xeMHQ8O3xvayWfflZHe1rJo3YkVJy3XXXceqpp7JkyZLS4+eddx4f/OAHmTNnDlu3buXLX/4y73rXu1i9ejV+v5/29nZ8Ph91dXVl52tpaaG9vb3itW699Va++tWvDns8Ho9PmNhJJpMAFSvspwJqj1OHg2Gfao8HDteVbOlMEc9a1ARM5jV5JQv3PvMGYWHRWGvQ2e+AEKVanZEQQHMQdOFi2Dka/Q6+kEajX2JnHWoNADkoNuRiCIFteo9bmRTC0ogGTPK2yyERmK3F+c//foDuV1dANk3A8Hq2fNF6Wo55F63zF9NY56WHupJ5nlm/nRqfRq1uEwl5Lc4untDxhJbEcSSGnaUhEiBSq5FMJPjbpl0saIns9f2arN/H8WRf9jh49tiemDRi51Of+hSvvvoqTz/9dNnjf/d3f1f6fMmSJRx//PHMmTOHP/7xj7zvfe8b8XxDK9kHc8MNN3DdddeVvk4kEsyaNYuamhpisfH3HygKqJqamin9wwpqj1OBg2Gfao8HhlWbuvjxys1sGTSNe15zhHMXt/JSW5aw30/Ohl2ZHFrBxM92RhY8xV21pWBHOo/rwuz6AG/2O+xOjvQq6fnkSNAFuLhoiTwLawXvbHb4/77/n6R2vzXwnmkGwfkn4Ft4MolIHa/3GywK6wgh6LNdXt+WJmhqZC2vTka6YLmFbjFDoAsNV0pc6SJNgd8w6Ui7JPFRU7P3mprJ+H0cb/Zlj6M9flKInWuvvZYHH3yQlStX7rWDatq0acyZM4eNGzcC0NraSj6fp7e3tyy609HRwbJlyyqew+/34/f7hz0uhJiwH6biuafqDyuoPU4lDoZ9qj3uX1Zt6uJL96+tMI07yYbdm8jmXWpDOkKAz9DJWC7aCBU7ovAfWRjnIAVoQiMa0DF0jXh/HlmQQoOnlhdx8YROa00Q4dp0de6mr+0NfvK7laR29wECAZjNcwkuPgOzYRa6LwASbBd6Mza1IZPetIUjIWiaZCxvqjoaSNcbJ2E5EmF4A7jcggiSjpd2awj7R/19mUzfx4lirHsc7fEHtNpJSsmnPvUpfve73/Hkk08yd+7cvb6mu7ub7du3M23aNACOO+44TNPkscceKx3T1tbG2rVrRxQ7CoXiwOK6kjU74qzY0MmaHfGSPb9iajJ0Grff1MjZLpbjUhMwyFkuGcshZzsIIWiKBgrjGYb/XAi84Zt+Q8PUBboQTKsJUBv0Mas+RFcyj+24A3U+cvjrpQRD1wjJFEamm7qAIC5ixFNeLacRjFB7/IU0nnYZvpYFaGag7DyW45aNdvAZmieqCucvluIUh4VK6YkuXYgDMgRTcYAjO9dccw333HMPDzzwANFotFRjU1NTQzAYJJlM8pWvfIX3v//9TJs2jTfffJMbb7yRxsZGLrnkktKxV155Jddffz0NDQ3U19fzuc99jqVLl5a6sxQKxeRh8AiAYipjaEuuYmoxeBp3Ku/Q2Z8jZzslEWBoXjlyVzLPzDqdiN9gRl2QjkSGdH6QH44AU9fQhVdQ7ErwGwLX1HGly4dOnM3tT2wknpXomjefaqhckoBrZQnYaSzhWZOYmoZe08yhx5xM4q0+okecTrCuFUcKxJA5WcDA8M/CaIeaoEFfRiNjeQ7Hxc4wCTiui5RetCqeHb1XjmJ8OaCRnTvuuIN4PM4ZZ5zBtGnTSh+//vWvAdB1nTVr1nDRRRdx6KGH8pGPfIRDDz2UZ599lmg0WjrPbbfdxsUXX8yll17KO97xDkKhEL///e+Vx45CMckYPAIg7DdojvoJ+41SS+6e7P4Vb1+K07jzjsvO3gxZy0ETAkMXaEKQtx2vRVxQGnMQMnVaa4KEfJ4/jl4Y1imQ2K5L3nHRgJqQSTxjM785wmUnzubaMxeWpn5rQpR1cknXwe7vJv3aU7Sv/DWuU0g3Fbq2Pv3xK5hx2vvRYq1evdAIEcfedJ7uVB5NQMink8o7xAImuvBSVUJAsUu8WHIU8uksmhbba9u5YmI4oJGdvXU+BYNBHnnkkb2eJxAIcPvtt3P77beP19IUCsU4MzSVUcy1BzSd1phWaMndzMnzGtRfvVOM4jTujkSuNPZBDKqp0XWB7Ugawz5aawJet1ZhzMGxc+o4fWEjD76yi3W7EuTsgfuGLSWdiRyG5i9FSy47cTYPr21n7a64N9tK15BS8mZbF6mdG0mtexIn0QEINrz8LLOWnETacll8yHTOW3YUbTLGNx56HcutvBevuNlzcg6YGt2pvGeAKPCGhAKOK3Ek+HTB4mk1nLukhVMXNB30LsYHkklRoKxQKKY+g1MZQ4sKhfCGH27uSJas+hVThyOmx2iOBWhPZDG0AaED3h+9jgt+QyeVs/mXcw9HE2LYmINF02Jc/cvV9GedgdfiRU5y9oAy0TTBJ8+Yz433rSGZc4j5XBKdu+h94TEyb70KhdJlISCbjNOelrS0tPLZC45F1zU+fvp82uJZ7lr1ZllhswDqwz6iAYNUzqE7lSPk09EKKTUhBFahVshvGNSHdK49cyGXnThbCZxJgLJjVCgU+4ViKsM3ggusX9ewXFlxcKLi7Y2mCc49wnPBd6QsREa8/1uuRBeC5qgfW3ozrIZOyXZdyTcffp1UzkHXvNELpiYwdYEmIJ6xuPWh9aVC92ULGvm3i46gycix5YUnWP/AHWTfegWtUCRsROqpOfkD1J14Mf6aJlprw5w8r6G03uWHNRMLmPgK59cKUZuc7SIEpPM2EmgI+5lVHyLoK8QNpFeUbBqC7156FH9/8hwldCYJKrKjUCj2C8VURt5xCWjD6+lyjluy6ldMPU5d0MhP/2KSybtYrlsaB+HTNVpq/Biahum4Fb//a3bGeaM96XViaVopMiiQGJqLQPJ6ez93r3qT4w+pZ3dPH9/99eP87YkHSe5+E9f1Ii8Bv49pS5dRf8RytHAtpq6DkGXDPwG296RJ5myQElPXSu3rWcthR28GpyDQTF0j6NMJ+3WyeRfbdb0UliupCaqf48mEEjsKhWK/cMT0GPObI6xv66c1ppWlsqSU9KUtFk2LqpbcKcoR02Msnl7DK9t7caXAlV6rlOU4Xt2NLjhyZi1HTI/hurJsYvdL23qxXder9RmSAnXxoimuI/nOI+vRcwkSPZ10P3U3Jg6GJnBcidk4h8hR59K48FBi4eDA611J3LVLEUXXlTy8ts2rzcFLXwm8+VtCh7zl4gJBUyNgelFKgSDo0wEd15XDhokqDjxK7CgUiv2CpgmuXu7VUrQnctSGTPy6Rs5x6UtbRPy6asmdBAwVGvtSVDv0XKctaODZzd04rlekrGmeUEnnHXRNcPrCRp7b0j3MmiBg6qU5WYPbqxwpsR1vHISbS2Pk0ySzFrbmxzfjCNydr+ILxQguPJXQ3GPAH6Y74xANDTjsD40ortuVYEtniuaon87+PJYrMbSCPw+UPgn59IqGdipCOTlRYkehUOw3li1o5JZLlpZuZsWOm0XTospnZxKwrx5Ig8XN9p40D69tZ0vnwLnSloNPFwjDc092XK9QOOTT0TWNB1/ZxS/+uo3UEJflzv4sUnqOxJomvcGdSGzHRToO+c6dIARGTcgbDKoLQgtOwvIHWPyOc9jthMjZEk3zipmzlkvQp1eMKBZry5qjfnyGTmd/lpztljyBgqZO1nbI2XLYWCIVoZy8KLGjUCj2K8sWNHLyvIZxix4oxoeiB9LwcQ6eB9Le/GGe3tjJdx7dwPaeNDnb8cY9CGiOBmiO+khkLRIZC10TzKgNYGial5rSvHRQxnJ4oz1JyKczsy5YZk0wozZIf64fx/XSSLrudUDlUwm63nie3leeIbrgOLS6U3AlaBL0UBj/MeejRWpplpKdvRmcgrqyHBcsKkYUB9eWRfwGYV+YrOWW1irxBI3P0FSE8m2EEjsKhWK/o2lCtZdPIvbVA+m/Vm7mu49tIG97dTjFRnAJdPbn8BkauuY1nLtS0pXMcUhDGCEGCtWlBNt1CfmGWxNomkZd0EdXKo8L2Nks+bYNJNc8TjLdhbQl2S0v0jntUNxADSIYQ/hDOEAyZ9EUDTCjLkh7PEvedohnLUKmXjGiWKm2rFiPI6WkPZFj8fQYV50+j5+s3KIilG8TlNhRKBSKg5x98UB6emMn331sAznLxdQFCMjbAyMWHNelsz9HS8xfEkqDU0lFMpbnnxMwh3fqJXM28YyFlC6iv5P+tX8mW/DMMQ0NXQOtaR5WIIZZ0wxCK12/O5Un6DMIF7qmDmuN8pmzFtIY9leMKI62tmzZgkaWzW9UEcq3CUrsKBQKxUFOd9Ib0WDownMG9mllxn9+XSNewQPJdSXfedSL6JiGQBeaN/aBAbHjAjnbEzJ+QyeTtxFCYLsu4AkbKSWZvIOpa2hDbJgkks7+HFYmSe7Nl3A2Po3IpjB14fk2RevwH3Ym/llLEIZZ9lqB52bcHs8S9nsztz5/7mF7jbyMtrZMRSjfPoxJ7PT2wp13wvr1XsHW4YfDP/4j1NeP9/IUCoVCMZGs2tTF95/YSCLj1dRomsBv6DRF/UT83i1ipA6jdbsSbOtOIYRAGzT+odi5VJww7kpPdDRF/WzvcXClxHYlritLEZOaoMGs+iBt8RyB2ECnUzqTp2/76yReeQy7ezt+w1NDDhrhw05h/gln0OaEC77I5QjhRabytsNhrdFRCZ0iqrZsalG12FmxAi66CGIxOP5477Hbb4d/+zd48EFYvny8l6hQKBSKseK6ktfaKt+wi0XJiUweQy9M6paSjOWwszfDjLogYZ8+YodRTzrvFQQzIG6KoxikHJgULgvipibopZKEEDiOS0cyVxYxAUrpo5qggZZLsbujg9TWV8h3bUMAWcvF13QI9Uefi9k4Ey2gQary3j3R5iORtfnMWQuHCZ2R2uyHt8w3KpHzNqdqsXPNNXDppXDHHVAcKu448MlPes+tXTveS1QoFArFWHhlex8/ff51NnekhrWSnzyvgTtWbKYnlcdxvUnibiEKI5C4lKd/KnUY1Yd8BE2NvK0Voj9eJMXQvE6uIq6E9r4MuxOChrCP73zwKGqCvooC7JZLlvLDx1/jjS07SGWypPI2wXnHkW/bgIYksuRdmHOORTN9yJKc8ih54eCNlLBdr2U8ZOo0hv1lx47UZn/6wkZWbuwac/u9YnJStdjZvBn+7/8GhA54n193Hfz3f4/n0hQKhUIxVlZt6uI//ryJrQmX2pB/WCv5x06bx2u74mQsB1eCoWloeBGYomDI7SX943UuRfnb9l6EA3nHRRODJYeHVgz3FJ7RhFfrUoyg/GVTF/UhH4taIzTJXi6qayd4waF877ENtMczWLqA499LsHEmeriuTEjtCdeVJDIWR86qLYtKjdRm/8r2OM9u7ibs12mOBqpuv1dMXqoWO8ce69XqHHZY+ePr18PRR4/TqhQKhUIxZlxX8uOVm0nnHVpjwVJ30uBW8l/99S0SWRtXytK8KU0X6Jos1dRoAq49c8GIN/jntnQTz+TJ5J3SrCtXlgudhsKkcL0wHbwrZfGtR97gOtflv/6ytRRBEVYa/a2/It58nohPcPnV1xPPWNRHfORFEGKtILRh0RwAQxMUut5LeINGwTQ0ls33am+KgqfYZt8S9ZOzJam8jS4EtuN4US5H4je9Iu3Rtt8rJjdVi51Pfxo+8xnYtAlOPtl77Lnn4D/+A77xDXj11YFjjzxyvJapUCgUitGybleCLR1J5kQNklKUiYBiK/nOvsJAS6183pQQwhMmuF5kJG0PO7/rSu55fhu3P7GRZM7yXi+HixCgMEhTsLvgROxKeHVHHx+7ezUBU6MpbED7Bt589vckOr2ZVDPrgvzp9w+QrVlGVsSwhYEU0it2doanrgxdIDSBZbulvdquJ4KkhJ89/Sb/8+xbzG+OcO4RrWzuSOI3NN7qSZfckcErotY0L0KVzQ+0xu+t/V4x+ala7Py//+f9//Ofr/xc8WdeCK+WR6FQKBQTw0gFtsWRB6auw3Ctgl/XvMJiIXAciRAuSFHqpEJ459Y0QV2ovJ171aYufvTUZl54s4ec7TI8aVVOeyKLXkhjGZpAQ5JzPUFhJTrpf24lfZteIu84uC44SNqDh1A7fTk5J4KQnmjRDa9w2BmUZpMSdAGW7SK0Ytu6wBACW0rCPp36cLkb9Ib2fpI5G8txcQrnFoJS+s5xQQpZ1hpffM+Gtt+P5xwxxcRStdjZunUilqFQKBSKatjTHKviyANrhNqWnOPiMzxPnFTeKZgAehKiWF6jCUEsYNIQ8Zdd88b71tCXtnBcF0PzIihFirf5oeLHkeDXvXPaUuJaeXJbnye5dgUyn4JC55YZqSN2zLn4Zx5Bh6sXiqW9lJRAoGsamiaxbBcE1ARMIgHQMnYhKiOI+nVcvJTZrLrQMDfo7T1p0nkHTQxEnQB0zRM8UBRR5aJlaPv9vs4RU+xfqhY7c+ZMxDIUCoVCMVr2NsfqaxcvYV5zhN7eOMJnloqDYWBYZWsswKbO5DCBUpwu7jcFi6fHSnUug0dKxAI6/VlrmKgZMpR8CAJXOmS72+h74QGc7m1I6UVTNM0gdvgp1Cw9E80fJG97Q0KL58w7EoH0IkOlyIngn88+lMUNJv2uSV/GpjZs0pey+PYjrxP2GxXdoKMBg3jWHphgXnrOK6Qu1h4Nfm7ogM99nSOm2P+M2UH5tddg2zbIlxtq8t737uuSFAqFYmqzL+mP0cyx+snKLfzTafO4/eFX2ZrIURPylY08CPs0hAC/oeE4nuAoprBcKXFcLxpz5amH8MDLu9jZl8Z1YePufgSwK57DGSF3NVJKy85lsFN9SDuPm+orFTL7mg4hcswFRJpno2ui1AJf6byWK9GkJGBomIbO7IYwC1p81NTUlN6HFRs6PWdlXRt+EkDXtJIfkO1IdG2g/GJw83o67xAw9GFjIoB9miOmODBULXa2bIFLLoE1awb/gAz84aDqdBQKhWJk9jX9Mdo5VjVBH9e8cwE/fb6dzR2pspEH5x7Ryo/+vInmaAA77I1jyNleR5UQgoDpiaFP/eplMjkblxHrj/eKdB2cVB+OnfXa0jWD0KLTSb32ZyJLziIw7zg03SDvuJgI7AoqanBdUNGs0NCgNmiyaXeS5O48DYVZV4Onlge04XO2HOl1mdVHfGTyLjnbQXrD0An5dPyGVqrpGWp6uGxBI2t2xMc8R0xx4Kha7HzmMzB3Ljz+OMybB88/D93dcP318J3vTMQSFQqFYmowHumPYvHxSJGLwYW0R8+q5eeLZ/NaW38pirSoNcov/7qNVN4hYHqOxmF/iGzexXZdDE2jP2uxuz8HeKJCjs7WZhh2fDf9rzxCcO6x6OG6UjTH3zqfwJwj0QORsuOtgtAxdYHjylKEp6ylHM9F2XElX75/DYaTZWdKYOga85sjXHX6vGFTy0uvLczgigZNpIQ5DUFylizt228KdifynHBIPf9y7mH0Zaxhkbdq3n/F5KFqsfPss/Dkk9DUBJrmfZx6Ktx6q9eW/re/TcQyFQqF4u3NaNJPo0l/7C1yMbSQdvCwylWbuvjo3S+wvi1Bf9YilbMImMagOVg6Ukq2dBWEjigvQB71Xq086ddXknr9aaSdw830Ez3+Iq9jKhhD84eHRUUGIxDootyzp+hVWEy5WY7kjd39HDctQHPUJOdI1rf1c9P9a7n8pNls70lXnFoeDRhcftJsfvnXbexO5KkNmYR9BjnHZXciT8Sv88kz5nPUrNpxef8Vk4PK0nQPOA5ECmK8sRF27fI+nzMH3nhjPJemUCgUU4fRpp/W7Urs8Tyea3GE3rSFHJJbKhbSzm+ODJtjVYwqvbYr7tW86BquC+mczc7eDN2pHP1Zi45EthRRsatMXUnpkmvbSM9jd5Bc+wTS9kST09+FtPMYNS3ogcgehQ54bellQoeB9JXA66IarAc1TRAwdVpjfpI5h5Ubu/jaxUtYNC1KOmfTkcyRztksmhbllkuW8vHT53PLJUtHfH5P0bWxvv+KA0vVkZ0lSzzjwHnz4KST4FvfAp8P/vM/vccUCoXiYGRvRcfjlf7QNMHVy+eXBmYOjVwUC2kHX7sYVSrOwUpkbVwXikGbvOOyq8/zwxmp8Hhv2Ok4yVceIfvmywxOPAVmLCZy7PkYkYaqzqdrAul6NU0UWskdCT5dw3Zc73np+e4UO6eG1izd/dETR/yejHWq+Vjef8WBp2qxc9NNkCpMmP3a1+A974HTToOGBvj1r8d7eQqFQjH5GU3R8VjTH5VE1LIFjdxyydLSNQcXHxevOTjqsG5XYsgcLOHNN3RkmbgZi85xHZvM5hdJrX0CN5csPa6H64gecwH+mYv3GsmphCYlAs/VWBRMDnVNoyZo0pXMlbrHdE2ULXywaBycwqt4jb08PxKjef8Vk4uqxc655w58Pm+e14Le0wN1dWVWDgqFQnFQMNqi42L6Y6TC2cE+LoPPvScRtafIhOvKUqfSlq4UiYyNy8AcLEfKYVGcSi3fe8LNJkms/j3ZbQNzgoRmEDp0GeEl70Iz/Xt49Z7JFyNPBeMfXUB92EfEb9CVzOI4EPJp+A2NlDXwuv1VMzPWyJDiwDBmn53B1NePx1kUCoXi7UW1RcfVpD9GK6IqRSY8kbSJZCLBzpQg70gsV2JolIROfiyVxwWkbeGke5G2ha95Xkns+JrmEj3uvZh1rWM+dxEXGKobelI5tMIeXClpjPgZ7P43kmicKMYaGVLsf6oWO5dcUjmCIwQEArBgAVx22fCp6AqFQjHVqKboeOnMmlGnP6oRUcV1FKML8Uyem+5fSypnsbBWpzlq0p3O05cpdla5pbEI1SKlxEn2IK1s6TGzYSbBQ47BbJlH4JBj0bSq+14qYuogpcDUvTERWcvGciRd/XkOa4nQmcyTzDnkghLX1ciqmhnFHqha7NTUwP33Q20tHHecVyH/t79BXx+cc45Xt/PNb8ITT8A73jHu61UoFIpJw1iKjkeT/hgqoqSUZK0BH5yaoMHmjiT3PL+NR9a1l9JchgYZ20VKyey6ID7DRZOCsM9E17zi5LEKHTvRTf/Lf8RNJ4iecHFJgAnTT+zE9yH0cUkUlNCFhiwM6Jxe60cQIJ33BM/XLzmSVM4uRa860t77ompmFCNR9U9na6sXufnhD6Eo4F3XMxuMRuHee+ETn4AvfAGefnq8l6tQKBSTh2qLjocWG5+2oLFiBGKwiErmbDr7s+RsFym9KHpRXN3+5EYcV5bSXImsRaI/hyYEqbxDg68QETI1AqZGKle9xb3r2KRff5rUa0+VWsnzu17HP2sJeqgGzRcsO14DIkGDRKbCuPVRUky3gUS6XpFyJKDjSp1UPs/L2/r48ClzOGluPX/btIskvpKDsoroKCpRtdi580545pkBoQPe59deC8uWwS23wKc+5XVoKRQKxVSmmqLjasZEFEVUXyZPZ38eR3pDMEWh8SiTd7ziXQFzGwYM+nRNoBUiQZ39OWYXJpYLIagJmFWLnXzHFhKr/4Dd11Z6TDMDCH8Yo6ZleOoOWDw9xu5EDoE9pu4uoCRYiuIu77i82ZUma9m4Em7/80Yef303nzh9Hke0RMpmYykUlag6uWrb8Prrwx9//fWBuViBwOg6s2699VZOOOEEotEozc3NXHzxxbwxyJnQsiy+8IUvsHTpUsLhMNOnT+cf/uEf2FV0MixwxhlnIIQo+/jQhz5U7dYUCoWiKopFxxG/Tnsi57V2u5KM5dCeyJXqR57b0s2N961hfVuCsN+gOeon7DdKxcarNnWVnXdRa5SmqJ/2RA7HdTE00Eq/3waOc13KpnMbmjfTStMEOdshN6gI2Wdoo/6F72STxJ/7LT1P/LRM6ARmLqHh/M8QPvSUiuKiOebn/k++g59+5HiOmB4jYFR3i9HwolCOC670RkIYmqCrP0cmbyMlBE2d2qBZcEtewyvb+6q6huLgpGqx8+EPw5VXwm23eWmqZ57xPr/ySviHf/COWbECjjhi7+dasWIF11xzDc899xyPPfYYtm1zzjnnkCoY+aTTaV566SW+/OUv89JLL/G73/2ODRs28N4Ko9U//vGP09bWVvr4yU9+Uu3WFAqFomqKRccjufGePK+hrNg4YOrDHH/vWLEZt1BLUxzp8GZXCsf12sPzjizMipLYjkTTBJoAy3HI5gcETcD0WrFdKXHdguFeAV2IUqRlpEyP67qkt7xI959+QGbraooGNnqkntrlH6H2tMvRQ7UVX6vhpZvWt/dz1Kxabjx/ES0xv+eDU4GhjwqgtSZASzSAJiBfsG8ebByoaxrNsQBBn1F67367ekfpvVMoRqLqNNZtt0FLi+ecvHu391hLC/zzP3t1OuAVKr/73Xs/18MPP1z29V133UVzczOrV6/m9NNPp6amhscee6zsmNtvv50TTzyRbdu2MXv27NLjoVCI1tZ9b3dUKBSKatlT0XE1U7L7s1ap3TzsN+jPeWkbKb1Ujq4JgqZOLGjQmch54sd1gYF6oVjAJGN54xYcV+LgEs85JDIWuuZ1ZJm6wJGUTRh3rRz59k0k/vp/A+vTDEKHn0p4yZn4TGOPc7K89JrNig0dAJw8r4GvX7KUmx9cx9auVJmHj6ENpKj8hk7A1JlWE6AnlSdtOcQCBk5h3/1ZC60wib0pGijM8Bp473b0plm3K8GRI8yyUihgDGJH1+FLX/I+EoURLrEhdgaDNEhVxONxAOr3YNwTj8e9H/La2rLHf/nLX/KLX/yClpYWzjvvPG6++Wai0WjFc+RyOXK5XOnrRGEjUsphs07Gg+J5J+LckwW1x6nDwbDPidijELBkRvkvQykl3akctuPi1wViSBWLRCJdl3Te5sU3u3ni9Q5SOYtpsQBZyy0U6nqDMW3HxadrzG3wCoL7UnlytoshQCDLCpldV6Ih6Unm2JHOIWVhfpShkco7WLZEKzgQS9fBTffj5tMY4VoCMxaR27kes3keNSdchBFrAsDUKIsUidIeBkjnHX66cjO/+utbNES8ouxMziJkal6BdXG6lZTomiAWMFk0PcYnTp8/TCwuao3yy+e38f3HNtAQ8RHy6aWi5SIBXeDYLt2p3JT9eVX/Hvf+2tGwT72CQ0XOviCl5LrrruPUU09lyZIlFY/JZrN88Ytf5LLLLiM26OKXX345c+fOpbW1lbVr13LDDTfwyiuvDIsKFbn11lv56le/OuzxeDw+YWInmfRs1KdqEZ3a49ThYNhntXt0XcmWzhTxrEVNwGReU3jUXT8R8swIS4K6hc8YiMBkLIfeVB7Hdojp8L+r3iBnu8yK+Ij5XPBJhCXI2RJdSFwEUjr4XK8WyIoILFcQII/uWMicRb0pwfTEUSSgE8DGkVAT8lEXMknmbDr6B4qU021bMENRREBCwLsd2CecQe7QI4nNOxIhvOnjPkMn7NPoGmxVXKBcfsD0mI4tJV1x74/I1oiPhXUB8gUfHFMTvGtxC0unx6gJ+krvZX9/gtlRmB31RFIq1c+SRpO5NYKg6eIzhr/fedshZEBE5Et/LE811L/HPVMMVuwNISeJXLzmmmv44x//yNNPP83MmTOHPW9ZFh/84AfZtm0bTz31VJnYGcrq1as5/vjjWb16Nccee+yw5ytFdmbNmkVfX98ezztWpJTE4/Ep3TGg9jh1OBj2OXiPUrJHz5tVm7r48crNbBnURTWvOcInTh+dn4vrSq74+fO83tZPa8yPEIJkYdK443p5oYCpEwsYtPfnMAQ0xQKeY7Lt0tGfLUsfaaLQXRU0uOq0eazc1MULW3uxHBddg4Bp0Bjx0Z3MUm867ExBwGfSFPWxvSeD7UqcdJzE6j+S27GW8OJ3EpjlFVlqgQhaMIIQXjmnLmBWQwjLlrTFM6XJ43vCb2oIwLJdEF5B8SGFjjEpJe2JHIdPi/LzK07cq2Cs9N4N/h7uTmQ5aWaQ2/5+GfoIXkdvdw62f49jETu1tbXE4/E93r/H1wVqjFx77bU8+OCDrFy5ckShc+mll7J161aefPLJvQqSY489FtM02bhxY0Wx4/f78fuHz2wpdnJNBIM7xaYqao9Th4Nhn0IInt3czY9XbhmxHXzVpi6+dP/aCiMbknzp/rWlkQ17QtcFVy9fwI33raEtkac2ZLA7kfPM/YRAFxpNsSC6EIikRd6V7OrLeVO9kbhuefLLkYCEnrTNr1fv5CMnH8Ib7SlMXRDyGQR8Gtm8S9aWaH4NXRdkLZedfTlytkvmjWdIrXsSt+CCnN74LL4Zh2PWTUPoJninRxcwsz5E2G+wNZHCkaLQ2u6JkJFqgnOWl6wyNQ1NE2RtScaSBH1eTq4m5GNzR4rX2vr3Omph+HtXPmIj6jf4wHGz0HVtyv+sHgz/Hseyx9Eef0ClsJSST33qU/zud7/jySefZO7cucOOKQqdjRs38vjjj9PQ0LDX865btw7Lspg2bdpELFuhUEwBXtnex033j9wO/vTGzlIXVUvMj5SQKrQ/t8R8w7qo9sTgjq142iZjOQjhFRvPqAsS8RvYrls6l4TCpO/hkZRSrYyEzZ0p/u2P60jnPTEW9OkIBJbjtcC7rsSRnmtyf/tWeh/9D/pf/lNJ6AgjQHjp2ZgNs0pCp3iN4o0nm3fJ2Q4a3rR0b/r4yHstrtcuVFZLSaGI2sOva1hDXKVH+94N7Xb72sVLOUoVJitGQdWRna1boYImGRPXXHMN99xzDw888ADRaJT29nYAampqCAaD2LbNBz7wAV566SX+8Ic/4DhO6Zj6+np8Ph+bN2/ml7/8Jeeffz6NjY289tprXH/99RxzzDG8Q82rUCgUFXBdyW9X7yi0gwcrzp76zqMb6Ehk8Rs6b3VnyNlOWQdRTbB87tXeWLagkRMPqedbj77BL559i7qwSW3QRNM0pJR0JXOeaaAcWGMlGVUo8S09Z7lg5R1292dorQmSyjl09Oe9bisJuXSK5CuPktnyIlIOalOftZTo8ReiBwYaOQRep5amCWzHMyZsCJvYrpc+swsRHXcP3oTFtUm8gmahCYxBLrRjmUo+UrebEEzZWh3F+FK12FmwAE4/3fPV+cAHPAPBsXLHHXcAningYO666y6uuOIKduzYwYMPPgjA0UcfXXbMn//8Z8444wx8Ph9PPPEEP/jBD0gmk8yaNYsLLriAm2++GV0fbt+uUCgU63Yl2NGb3mM7+LbuFHlHkrMcL62jCUShZTprOeQsh5DfGHWEouig/HpbP1nbYXfCJZ6xaIoG0IUgZ7uYuoZbmF+1p4BRpae6khbJrI3tDnRNJba8QveqP+Fk+0vHGZEGosdfhH/awmHn0DSvJV1IT9xkLZuO/gGBZJT8fSovrvRWFp62JYR1r20c9m0qeaUJ45Ok5FTxNqBqsfPKK/Czn8H113tjIf7u7zzhc+KJ1V98bz+ohxxyyF6PmTVrFitWrKj+4gqF4qClJ53HdiQ+c+QBnq6U5GwHV0pMQ0MUEkhCgNC9Atys5VAbNCueYzCrNnWV/HNqQybpvE3WcshYDjt60kSDhidQNE/k+HQYwxgrskUjPsfGTfWR2LYetyB0hGYSXnw64SPeiagwxwugLmiQtd3SHC7HBb8hvLRVIaplCA1NyDJ35hLFsNMggj7DE4i2o6aSKw4YVdfsLFkC3/se7NwJd90F7e1w6qmeY/L3vgednROxTIVCoRg/6kM+jMIAz0rkCgZ+JYb8zSWll2JyXcmmjuQe63ZcV5Y5KAdNg+ZYAF3TkC5YrqQ3ZeFKsBy8NNEYAxZSSpxMP3Z8N9LO0XTEKQhfEF/LAhrO+zSRpWeNKHQAIgGTQxrCzKkP0xT1UxM0CPp0WmMBNE2Qtz0TQ0G5ptHFgCtz8e9TrfC468phrtJqKrlifzPmbizDgEsugfPPhx/9CG64AT73Oe//f/d38M1vgqoPVigUk5EjpseYWRfiuR0ZWmI6Q1ua+9IW02qC2E6GjGVjudIz+APswsiG4rFf/+Nr3P/yzooDPcFLmQ11UI74DerDPnYnvELhodrG2oNT8Ujk2jdi97Xja55XeswIRmk4+5No0Ya9dq3oopCqK7gV92UkM+tCbO/NkMlbBSdnieWAjSyrL9KEQNM8kea4El0ImmJ+kPAv5x5OfcRXsa1fodhfjLkb68UX4ZOf9ATN977nCZ3Nm+HJJ72oz0UXjecyFQqFYvzQNMEHjpu5xwGeHzpxNmG/TlPUT9D00lqWMyB0dM2rYdnTQE/wUmaWI/ENamGSUpLMWWiaVxCsC0FD2IehDfVY3jtOtp++Z+6l988/I7nmcdxcymtpD9WgR+oxYnsXOuC1tLfHs/Sk8qX34MS59SRzXsrN0AR+U/MKmAvCyChOJ0dSDJKFfF6HWd6WzG+OcNHR01l+aBNLZ9YooaM4YFQtdr73PVi6FJYtg1274L//G956C772Na9L6x3vgJ/8BF56aSKWq1AoFHvGdSVrdsRZsaGTNTviI6aYjppVy9cuHnmA52UnzmZ+c4ScLZlTH2J2fQjTKDgK6wDezb8maFYc6FmkPuTDHJIyy1peXYyhaaWoiM/QqhI6UkpSG56l+w+3kd32iveYnSe/awNGTQtaIFzV+6YJSFsO7Yks02r8fO3iJWzY3e+lpwqjJQReZ5VpaIVuMYnf0Aj7DJqifmbVhWiJBUjmHFWbo5hUVJ3GuuMO+Md/hI9+FEaauzl7Ntx5574uTaFQKKqj2PE0kkngUJYtaOSU+Y0jOihfvXw+N963ht39eYKmhu24pY4lXQiaooFS1GTwQM/BXUNHTI8xvznC+rZ+WmOe+Z3tegXACC8iomuwO5Edda2O1buLxPO/w+rZWXpMM4OEjzyb0MKTCg7I1cWI6sM+wj6DeMaiJmgSDZhs6UzRHPXT2Z8vS+UNnpB12Umz2dSRZHNHkv6cjakJFk2LjvieKxQHgqrFzsaNez/G54OPfGQsy1EoFIqxMbjjqeh2nLMd1uyI88+/fplrz1zIZSfOHhZpqNTSXKRoaHfHis28tiuBK7354sEhE7jB6+CKVzDL0zRREk3tiRw1QQPL8aaSu7bX4ZSzR7dHaefpf+URMhufG+SZIwjMOYroMeejBysPPx4NrpTEgiamobGlM8XL2/qwHElz1I/P0EtDRoteQwFTx9A1zjismS9fsHiPIzcUigNN1WLn1VcrPy6E57kzezZUmMSgUCgUE8bQjqfi/KniDTqelfzbH17j4bXtfPKM+Zwyf+9O7EWKhnYPvLyLf/vDOsJ+g5qgOawOZk9meUXRdOtD63mjPYnluAPRkdFGc7q20ff0L3EyA4MP9WgjseMvxt86f9T7GYl0zvHSUgXRJgWl9FvEbxD2hclaLrbrpd8kkkzeoT7k26NgVCgmA1WLnaOPHmQcVQHT9LqxfvKTfTMcVCgUitEytOOpNGhTSgxNYGrguC5rd8W58b41fP3iJRzRtHd/HPCE1LpdCWrDJrMbwuzoTVMzxFtnb2Z5rivZ0pWiI5HDb2hEAjr9Ga/La29I6eJm+nEdG+l45jtCNwkvOp3w4jMQ+si/xk1N0Frjp6M/xwhd9iXyjks27w3vNDXBMbNqh6Xfgj4d0EsDPfe0XxXpUUwmqhY7990HX/gC/Mu/eEaCUsILL8B3vws33wy2DV/8Itx0E3znOxOxZIVCoShncMeTlJLO/iyOlJiFVmopvBEHNQGDZM7hxys384NLDtvreYfWALnSJZV32N6bpikaKBtKObQgt3jDf3pTJ4+sbWftzji2LNS85IZ571XEzWdw0nFwHTTTT+jwd5DveJPY8e/FiO45OmXqXmdXxGcgI4L2Qpv7iNeSkMjmcaVXc7N0Rk1Z+m3oEM6RCpCrrZtSKPYHVYudr38dfvADOPfcgceOPBJmzoQvfxmefx7CYc9hWYkdhUKxPxjc8SQlhU6ngQnKxToTU9epDels6UiypTPFcXW1pXMMjUbEM3luqjDx3HJy5GyH3lQeTRMVC3KLN/zXdiXoS3tzqorIIf+vRG73FpJ/+xPhI89GD0QAEKaf0GGnetGcvbSSe3v3Nm67XhqqvLB4gMGP96QsZtQGSiJmcM3S5o4kcVfusQC5Ut2UNyXea81XhoKKA0XVYmfNGpgzZ/jjc+Z4z4GX6mpr28eVKRQKxSgZ3PEU9uklcQNFDxhJwNQJ+DzX4oQriWet0uuHRiMMDTK2i5SSWXWhskGhs+uDtMWzzKoP8dkzF9IQ8ZelaQbf8NN5h71MvCnDzaXof+mPZN58GZCkX3+a6LEXoIdq0HzBUZ/Hdr09OxJcx6XWb6BrYLvF+Vai9L4MXp+uCT522jyiAZMVGzqpD/k4eV5DxSGcQyM6leqmiu9ZcbjqHSs2c/K8BpXSUux3qhY7hx8O3/gG/Od/el1XAJblPXb44d7XO3dCS8t4LlOhUChGZnDHU1/aEzEuElFw9NWEoCnqRyDIOg6mJqgJeHU3laIRiaxFoj+HJgSpvFPWdVUcFLqjN81b3WkaIgMdGYNv+DUBk750flT1x1JKMptfJPnKQ7j5zMDjVhY9XIdmDu/6GOxgPOJ5gY5EjmTOIRowiWcspPQmkWvC+7Adzw25Meojk3f59QvbueOpzVWnoCo5RQ+sVYzYmq9Q7A+qFjv/8R/w3vd6aasjj/T+wb36KjgO/OEP3jFbtnjuygqFQjHRFNNPliv52GnzeGjNLl54sw/LdtE1ScD0DO8ifmNQIXGEeU1hXFfyo6c205e2qAl6AyuF5kU4NCEK9T85wn69NAg0mbPpSGTJWA7fe3wDYZ9eEgTRgFm64VuFlNrexI7V107i+fuwureVHtN8QSJHnkNwwYkFz5xyRhsX0fDEnhCC684+lO8/vpF03vF8flzv93fQ57lE522XZC7PjkI9UrUpqEpO0YMZqTVfodgfVC12li2DN9+EX/wCNmzw/rL4wAfgsssgWrB4+PCHx3mVCoVCUYFKxbDzmiJcfvJs/rSmjZzt0hjx4dd1MtbA1O1PnO7VpPzq+W288GYPjuuSzNkIAX5DJxY0vOnmQpCzHbJ5l6BPH+jycl00AQ1hH7omSoLg706YNahQmqEOfGVIO0//q4+R2fAsUhZHnHueObFjL0Ar1OqMyF5UVHH9PkOgAUfOrGXx9Bjr2xLUBAKFTjXNS+1JyY7eNJqAGbVBNM0TLNWkoAbXTQUqDBvdU2u+QjHRVCV2LAsOO8yL4HziExO1JIVCodg7IxXDvt7ez47eNB87dS4rN3YVCmvLnX1Pmd/AyrVv8cMnN5Iv3IQ1TSAlZPI22byN0ASO64VmPMdjrdDl5fVwB0yDUCHiUxQEj6zbPXDD92n4dJ2M6wxbu+vY9D7xX1g9O0qPGbEmosdfhL9luGeOBgzuHB9VGZAETfPW3pexWLW5u5Tqi2ftUndV1nLpSuZwJbTG/CWhU2S0KahKTtGlpeylNV+hmGiqEjumCbncnn12FAqFYqIZTTHsyo1d3PWRE1jf3j+ssNZxXH67egc52yl1bQkErpRIvKGYg1uoUjkbTQiylpf+0TRBLDi8jmd3PENLTZAdvRkifn2Qy/GgtWeTOJkEgdlLsXp2eJ45R7yT8OGnjeiZU+0Q9GJASRZb3SU8sq6Nq06fV7G7akZtkB09GWqDlaMuo0lBDXWKHm2rukKxP6g6jXXttfDNb8JPfwpG1a9WKBSKfWekYlgpJVnLxdQF69sSrGtLcNSs2oqv39GbpjHix0rkyFguGhLblWVRk+KZe9IWiayNXTABFBI6+3MkMnapHsinCbpsl4XNEbb1pHirO4cminUzDm42hbSzSNsroDZbFxJenCU4/3iMSP24vTdiUOrMdiW26325vq2fe57fxt+fPIeT5zWwZmecl7f1IQVE/QbfeeT1fU5BVduqrlDsL6qWK3/9KzzxBDz6qDf9PDxksO7vfjdeS1MoFAcre3PgrVQMO3g8hJTerKcb71vDl85fNOwm25POYzsSv6nTFA2woydd5mZc1Au6JphZF6ArmSdrefEVUxfohZRX1nLY2ZuhPuwjnrHI2w6Pr99NsjDsSgKZzjfpf+F+NF+I6LEXAKD5w2ihGP66aVWO66xMWWlQoShaDHpO1zyH5Nuf2Mi8Ru+X9tBap4ztkkrkmF0f3KcUVHG8hnJQVkwmqhY7tbXw/vdPwEoUCoWC0TnwDi2GHToeohif2dmXqdhJVB/yYeiCpO2gC41o0KA3ZZUZ/gmgOeYnEjDZncjjSknA0LBcF+ECeKLHclx2F9yJA6ZObcgkkbGQuTR9Lz9EdstqQGIDVvc2gnOPRxhm6TrVYmogNEHelqWvHTkoZTXoWCE8wYYQGHhmi7c+tJ5ExiKVd8pqnVJ5zxdoW0+G5ph/n1JQalaWYrJRtdi5666JWIZCoVCM3oF3cDFsS1SUjYdAgON6BcQzagPsTuSHdRIdMT1GTdBk3fY0Eq8QuSgSjELUJujTqA/7yOZdLMcpzYbKpV0cryKmbO2GgJqgya6+DP2bV5N8+SHcXAoAgcDfMAuzaW5J6IwV2wVTCHwFsedIgSbkMM8dAYXiam9vuvD2tGF3kqCpMXOQWaJf02gM+7GdLFJKkpk8cSq7QysUb0fGVHVj2/DUU7B580DL+a5dEItBZC/dkgqFQlGJah14i8WwO/uyZC0XXfPkh+MMmAhqQqvYSfTclm5603mv2wrQhDcbCjxxoGuCWGHYp9eJBQjoz9ojmvlJIWjbsY2+5+8n1/kmRTGk+UOeZ878Eyp65lSLBCzHG4ehawKfrmE5wzu+hsoxCXT2e3tuCA/UOnnpP28Ehlt4E6bVRrjkmJmcWhCWKgWleLtTtdh56y1497th2zavM+vssz2x861vQTYLP/7xRCxToVBMdap14C0Ww97yp/W81pZAOl6rdcDUS0XDMLyTyHUlP165GceF2fUhOpMWWcsuu57jylIBcjSgF8RNMULimQ2WCQk7T9/aJ0i/8QzCdRF47oSBOUcROeb80nyr8UIClisJ+3TmNoaJZyz6MhYBQ9CZtMqOFYCpa2jCq9uRgN/wRFcx/edKT+Bpuhc5ao/n+PUL2zhqZs2kFDrFmq7uVI4IeY6JxtD1ybdOxeSharHzmc/A8cfDK69Aw6Chu5dcAh/72HguTaFQHEyMxYF32YJGvn7JUj5294uYuiDkMwj4tJLbMQzvJFq3K8GWjiRzogaaNAkHTHqSedoTWQbVKHsjHPI26bzttau7EkOnLOVVPC67fR2p9SsLjwjMmhYix70XX8u88Xp7Sgyuy6kJedGnRNbCctyyfZfWV3zRoKcsx0XiCTpXSgy92HoPmpA0RXzEs3ZFI8G9FY9PNINrumzHZUZYEont4OrlC1SqTTEiVYudp5+GZ54ZmItVZM4cbyaWQqFQjIWxOvAunVHD4dOirG/rp94sFzqVOomKosrUdbAB6YkFIQSmAEdK3ELBb/EmXh/20ZnI4TgDnjcCcPJZnHQfRk0zvuZ5WD07iSxeTt2S03GFjl2tQc4IDBY4xf9rhXRae8LrQPPpGnUhH6mcV6hdfB144kYXAl0INAGJrIPfcMjZDrrmCR0pvdb7oKkR9OkITQxL/42meHwiGVrT5dcFQd3idTVVXbEXqk4gu643B2soO3YMjItQKBSKaikWHfemvWGVgymKlvnNkWHtz8X6nYhfpz2RI2N5tScZy6E9kRvWSVQUVamcTX/Woi9tkbO9GhhD17z/a9AcCzCnPszs+hCuC9GAgaFrSNch89arWP3dyFQ3unQRukH0hItpOO/ThI94J67QS1Gi8Yh5BEyd6bUBWmMBgqZOIQuF5bjMrAsR9um0xAJeukrDi9SIQQJJeqms5liAmqBJ2K97rsmuV9njSonlSnQhaIp69VJ+XcMaFEkrCo31bQnCfoPmqJ+w3ygVj6/a1DUOOx2ZoTVdAVNH0wQ+Q6c15ieZc7hjxeZS3ZFCMZiqxc7ZZ8P3vz/wtRCQTMLNN8P554/jyhQKxUFFtaJlMMX6nUXToqRzNh3JHOmczaJp0WF/7cczedKWQ0cyx/aeNG3xDJYjsR0XtxDVCZgGDREfQZ9OwPCiTLPqQ2h92+h99D9IrLoXZ+c6TEPHH63BrGnGjDWVzAEdd6DgeSy3Xo2BzJMmoCXmpyHspynqZ15TiEjAZNG0GHd+5AQ+c9ZCNKHh0zUCpoaueZPMhxZRRwM6Odtl8fQY3/7Akcxt8uqIbNfzJAqaGjPqgqVap8GRtJGERsDcf0KjmpouhWIoVaexbrsN3vlOWLzYK0i+7DLYuBEaG+FXv5qIJSoUioOFPTnwXnX6PKIBkxUbOivWigw2s+tK5ehLWdSFTKIBE9eVaJpg1aYubrp/rfe17vnxFKdCOBIc20UIiPiNUjos57gIK01s05/Z/chDWJYDSJIbniVy+DKkZmIICJka8awX9haAoYNVIQq+N0zdSzdZjltKpzmuxHVlyfemNmjypfMXcdSsWtbsiJfSf7YrsZ3KgqMrmac56i+lnE6e28AHfvIsWzpTNBWEXVFEDE3/VVs8PhGoqeqKfaFqsTN9Orz8sidsXnrJS2tdeSVcfjkEgxOwQoVCcVBRyYE3nsnzk5Vb9lgrUiycfXpTJ4+s283ueKZUM9Mc8/OhE2bxyLp2kjmb2fUhdDvLtv4KN0YJXckcmga1AYO3XllFbu2jhGMwrzHEW70ZrOh0ose9F0cYnhqRlIQOAAKkFOiat67RxDuKdTkCb1aXpnklA6Yu6M9apPM2fkMf5nsz4DmUIJP31mDqoizKIqCQovJz8jyvs8QwND5/7mGlwaBCEyMaCU4GoaGmqiv2hTH57ASD8I//6H0oFIqDi/3RjTPYgbcYjRluNJjg+t+8wj+ccgjRgMHDa9tZ35agL51H4tWoCDyfnK5kjpsfXIcmBC0xPwDpvOeZ49O8LqTi3CuJ9/n2t7bx2uoHcLvfYmFTGCF8zGpt5OYbr6SvcQk/e+ZNtvWkS+3oPl0UupxAE4KaoFlq8e5JD4ybgEKKSvP+WCwiBz2ZdxycwnOWI0lkbYKmznuPnsGXL1hc9n4X03/X/+YVeqw8ulYoRNY8s0FDEzQVan26k/my6MtoZ1lNBqGhpqor9oUxiZ0NGzxTwY6O8n+sAP/f/zcOq1IoFJOSsXTjVCuOXFeWhlQ6SB58eecwo0HbkmTyDj1Wnu88+gay6BNTOK0mBLlCWMfUBD7DEw22K+lIZEFKTFd6gkhohQKbgtiRkuSrj5J+42lwHExdoOk6Z599NldddRX19fW4ruTR13bTl85TE/Rh6hqW47KzL4OGJ5Z6Up7wEAL8hk59SPeKrymMcCj69ugCDU8oAVi2WxYJMgv+Mem8w/88+xbTawJ8/PT5Ze/ZyfMaOGdxC7/467ZC6kuiCQgYGrPqg4T9XiovkbWHRV9GM8tqMgiNSlPVA7ogZzu0JywifkNNVVeMSNVi57/+C66+2qvRaW0tTNgtIIQSOwrFVGW0oxyGvqYacbRqUxe3PrSeDbuTXpSkMNTS0AQ+wxvfkMo5ZUZ4xRoV25GFYym1XoP3uZADDem2Cz2pPM0BOajl2hMaUkp0K4XbuxMcB7+pYcRaaD7rcr7whb8v3UjX7Up4tS5Rr1gXgPxAVKiIpnnXzeRtMkDQ1PAZGo4LsaBBX9oiazk4eMKrIWyyuz9XKi7WCrOtBAJNuORtyX88tZmPLpuLUYgaFd/j9W0JrxZJeCmqhrCP6bU6PZaBZCD6Uhs0WbMjPkzY7KnWppLQ2JfZWWNlaCQq4bqYIcnh02LKZ0exR6oWO1/7Gnz96/CFL0zEchQKxWSk2lEOUL04WrWpi3/+35fp7M+VupDsQSMcdvVl6UnlEVAywrMdd1g9zFBvG1dCflDRrgSyhehJ8cOVnguym+5Dxya66FSsvnZmHPtOWo4+g15HlKV/KtWw+I0KcyTkoPMDGcsl6DNASLKWg6FrYLtoApqifrw2cO+lXpHzgG+QJjQM3aU/Y/H7V9u45NgZZe9xbdAkk3fIWg6W7dLVn2dm2EvZFaMv02r8fPuRN9jSWb1PzmhTXhPN4EhUyUF5wXT0EeqJFAoYQ+t5by988IPjc/Fbb72VE044gWg0SnNzMxdffDFvvPFG2TFSSr7yla8wffp0gsEgZ5xxBuvWrSs7JpfLce2119LY2Eg4HOa9730vO3bsGJ9FKhSKqtt+996qbPOtR97gz290sGZHHNt2+dFTm0tiRte0Us3KYLKWS6ZQ++K6w4VNNbgScnmL3nUrSW98DjvRie7aCMOH0TSX2e//InNPPhc0g1Te4aW3ektFv4NrWIrkbDnMVMeVbmlEg/deQcinFyJVOp9YPp+vvvcITprbgJSS3rQ36kEApqGhD3mvtYJ3zs6+9LD3OOgzaI4F0DWtMAzVpSeVI2PZtCdy6Bp09Od4vX3sPjnLFjRy90dP5CcfPp7vfPAofvLh47n7oyfu94hKMRK1/NAmFrREVOpKsVeqFjsf/CA8+uj4XHzFihVcc801PPfcczz22GPYts0555xDKpUqHfOtb32L733ve/zwhz/khRdeoLW1lbPPPpv+/v7SMZ/97Ge57777uPfee3n66adJJpO85z3vwankfqhQKKpmNN04gw3o9iSOUnmHVM5hzY4+/vnel7nqf17kAz95ljU740gpMXTNG765h/XYriwTGmMh3bGNrkfvIPHSn0i++hiunccNxJDhRnTTRywc5M3uNNt6UvRnLW7/80Y+ctfzrNrUVdEA0Xa9cQ3moLeoKMYE4NM1NOEN72yNBbAcyarNXVx24mz++x89AfHhk+Z4aahC+/lQ3IKemlEbqvgeR/wGM+qCBE2vjTxnu8TTNoe3RmiO+nFcuc8+OYOFxtJJOjtLoRhK1WmsBQvgy1+G556DpUvBNMuf//SnR3+uhx9+uOzru+66i+bmZlavXs3pp5+OlJLvf//7fOlLX+J973sfAHfffTctLS3cc889XHXVVcTjce68807+53/+h7POOguAX/ziF8yaNYvHH3+cc889t9otKhSKIVTbjTOSOCoOnrQdFwQETI2QX2dLZ4pkzgJZPoF88JiE8cLNpUm9+ih9b72IbTneNew8bjaJ1jzXu7Yr2d2fK60naOrUBs2yFNzQGhatIDhc6RVF14RMetNWqe5GFoSKoWkVvWmWzqxhUWuUX6/eTjxtoQkXbdCUdFe62I6kJmRy4ZHTeGZLd8X3OOI3CPvCZPI2ppvjH9+xgGNm13H1L1YfUJ8cheJAUrXY+c//hEgEVqzwPgYjRHViZyjxeByA+nrPhXTr1q20t7dzzjnnlI7x+/0sX76cVatWcdVVV7F69Wosyyo7Zvr06SxZsoRVq1ZVFDu5XI5cLlf6OpHwQu9SymE29eNB8bwTce7Jgtrj1KHSPhdPizK/Oczrbf20xvzDunHi6TyHT4uyeFoUKSV1QROfDpbjlMSRRLI7nh6os5HQncySyhlEAwbJnFdgbDtOWTZovOIGUkoyb/6N5MuP4Gb7CRgaCI1Y6xz8R12AXjedgY6sgdcV29VDPp2gWaxP2sTPrziRr1+8hB+v3MyWjiR5x0UXIAXMqA0AEE/nMYQn2RxXEjB1gj6BQBLQBQnXpTuVK73Xui64Zvk8vvXIG1i2i6HLkvizHa+N/Jrl89B1UfE9Lr1nAnQNQrrgmFm19Kbz2I6LX/euPZRKa3k7cDD8m1R73PtrR0PVYmfr1qrXMiqklFx33XWceuqpLFmyBID29nYAWlpayo5taWnhrbfeKh3j8/moq6sbdkzx9UO59dZb+epXvzrs8Xg8PmFiJ5lMAgz7q2qqoPb49sR1JVs6U8SzFjUBk3lN4cIImOH7/NiJrfzHn5Ok8xkiAaNUcNyfs5kb0/nYia3093t/OMwMS46dFmBrV4p60wcI4tk89T4XClYsmvD8aaR0wHGYHRbYE/QLPdvXQdtzvye/eyu+QstWOBxi2pFnUb/weNBGzuj7DZgVAfDS4pFajWQiwd827eKIlgg/uOSw0nu4O5Hlvr/tJJO38BsaMgQIrzpZCGiOmQRNL7eVsx3MkCRCvvSHHsAHj2rEcLL87+odpHN2wWgQQn6DS4+bySVHNRKPxyu+xwNIunN5Dm/xMzMssTJ5ZoQlQd3CZ1SIzI2wlsnOVPw3ORS1xz1TDFbsjTH57EwEn/rUp3j11Vd5+umnhz03dPNSyr2+IXs65oYbbuC6664rfZ1IJJg1axY1NTXEYuPvE1EUUDU1NVP6hxXUHiczQ/1u4pk8//mXLWwZ1BY+rznCVafN44imyLB9nl5TgxEIlyIZlmtjaoJ5zTV84vTh3Tgfesdh3HT/GnZ32dSGDLb3WORs7zmBZ/qna97IBst2Sx1L44mUkuy2NST++n9Ixyv+RdMJzjmKmaecQ4eMsjOz5++lroHmEwR9nkhwXY2OtEsSHzU1XsrnuLra0vFzWhv58crNbN7dT1fGG/UQNHWaYwEyGGQsb13tCYvDp8U4ZsH0YXUvV7yzhr8/bRF/eLWNnX1pZtSGeM+R00rt5kXK3+PB7eA2Ub/BOUfNoa6ultraWiKxHYXInDksMrentUxm3u7/JkeD2uOeGe3xoxY7ixfD009DIcPEP/2T14Le1OR93dEBhxwC6XRV6wTg2muv5cEHH2TlypXMnDmz9HhrayvgRW+mTZtWeryjo6MU7WltbSWfz9Pb21sW3eno6GDZsmUVr+f3+/H7/cMeF0JM2A9T8dxT9YcV1B4nM0P9blwpSeVtfLpGSywwqC08yZcfWMtX3j2P02trh+3zHQubOGV+46hMAk+Z38iVp83n3ue3saM3Q9YGCo3Uhq6haQJXSmRh5OW+lRtXxsmlEf4wUngpKz3WROy49+JvmYsRApmicP2RsV2IZywkgoBPI+tIDE2jIeyv+HMw+D16elMX//3sm1iOi1boMBvwpjG4evmCEVumTVPnkuNmVnxu8LW+fsmRpe9tX0GAHj4txidOn8cRTWZh9ITg6uULuPG+NbQl8hV8cva8lsnM2/XfZDWoPe75daNh1GLn9dfBtge+vvde+OIXB8SOlN5g0GqQUnLttddy33338dRTTzF37tyy5+fOnUtrayuPPfYYxxxzDAD5fJ4VK1bwzW9+E4DjjjsO0zR57LHHuPTSSwFoa2tj7dq1fOtb36puQQrFFGSo342pC7Z2pclZLq7ruQoHTFHyzNmdyPLb1Ts4dfFsdH34L5K9GdAVr1kmrgBdePUvTsHd13E9Mz85JKJjaPvWUi5dB6SLk4oj7RyaL0hk6VlIxyJ8+KkI3aTaGFJXMk9vxsKnaxi6YG5jhK5UjjU74hXFXvE9WjqzhqNm1kyoN81IDshCUJaSqtYnZ3+MBVEo9hdjTmNVSq1XKzqvueYa7rnnHh544AGi0WipxqampoZgMIgQgs9+9rPccsstLFy4kIULF3LLLbcQCoW47LLLSsdeeeWVXH/99TQ0NFBfX8/nPvc5li5dWurOUigOViqZAWbyDrbrYuoCR0Jnf5awL1z6q6o2ZLKjN826XQmOnFVb9U2vkplgImuRzFoUG8pz9sit5WMVOtJ1SG98jvSGZ4mdcAmaL4Aw/eihWsz6GWM76eDzS0kq7xVPb+ns5/O/eXVUpnyjGcewr1QSoJXqD0e7lrGMBVEoJjMHtGbnjjvuAOCMM84oe/yuu+7iiiuuAODzn/88mUyGT37yk/T29nLSSSfx6KOPEo1GS8ffdtttGIbBpZdeSiaT4cwzz+TnP/85uj68EE+hOJio5MVSjKboukBIT3hkLbdUk+LXNWzb88yp9qZXSVxJJL6CQZ41Ch+XsWB1bSfx4gPYfW1I6ZLZ9FdiJ16C5guOy/ldBuYAeh1jMLPWh+XKPY7LKDKaaNj+Ym9rGctYEIVisjNqsSPE8MjNvqYPR9P5JITgK1/5Cl/5yldGPCYQCHD77bdz++2379uCFIopRiW/G8/nhZLvi5RFAaSRtVzSeYt6E7Z3p7jzmTeruukNFVfJnE1nf46M5eBMgNBxc2mSrz5KZsuLXvoKELqJFq5FmIFxu46pwaCh5eQch7wjCfpGHpfxdmQsY0EUircDoxY7UsKZZ4JReEUmAxdeCL5CG+ngeh6FQjE5qGQGGPBp+A2drOWgad4fLXnH5c3uFDnbxXFd/BHB9x7bgAvMqguN+qZXFFemLuhO5ehIeEMt9+yHXD1SSrJvvULyb3/CySYBiRAaZsMsosddOOa0laGB45ZX9PgMrfCH2aDhoq7nbxNEn1KmfNWMBXk771Nx8DFqsXPzzeVfX3TR8GPe//59XY5CoRhPimMN1rf10xrznHsFgqaonx09aSxbYhoanYlsybXYEIKI32Bb0kIIjVTeIeIf+FWxp5tefciHKyVbu1Jkrb1LnLE4JNuJDvpf/D35ji1I6QICzR8mcuTZBOefgBBj7yhyhhRL6wKsEeqLMnkXISwMTcOnC+KDxmVUy2QpBh7NWJB92adCcaAYs9hRKBSTH6/luHysgb/gbxPy6+RsbwSB43p+MgHToDnqI+hz0ISNKyWd/TnCfr00fRtGvunFM3lSeZvcKIQOVC90XDtP75N34mQ8IzGh6QRmH0Xk6HejB6N7eXVlAoagPuwnlbPRNcHRs2t56vVOXDzxMxJdqRxaWiCElxoM+bTSuIxqmEzFwNWOBVEo3i68/UwVFApFVRRbjhdNi5LO2XQkc6RzNkfOrOXGdy8iEjCoC5u01gQ5pCFE2KfjuBKJRAjIWjbZfHmLVKWbnutKfrJyCz5d2+d6vkq4+QxOsofA3OMQQsOoaaHujI9Sc8oHxyR0BBA0NeY3RWiI+JlVH8J24dnN3aPy/HGlJxC998ghlXeIZ6qLeBSLgde3jX0S+XhSacBpESklfWmL+c0Rjpg+/uarCsVEMmkclBUKxcRRqeU4nsnzjYdeJ5GxEEAiY9HRnwUpmR6UZbUr/Tmr1K1VvOktmhYtu+kV6z0ifoNEdnyK+Jx0HISGtHNIKwdCEDz0FPRIPcFDjip45owNXQhaa4JohVERqbxDOm9V1fruSq+e0dAEfkPjJyu3sGx+46hSUJOxGHikSOCA+aDO1cvnq+JkxdsOJXYUioOEwS3HqzZ1cdP9a+lL5wvTuj1xY7sSgcQekr7ZncghgVjAHPGm15POk7OccRE6nmfOX0mtfQKzaQ6RJWcizAB6uBah6Rjzjx/TeQfXCOk6hAYJuM7+bNXT1l3pTURvivrRNVFV8e5kLQau1nxQoXg7oMSOQvE2ZaxFra4r+dFTm+hL56kJmqTzDnln79frSHjprwXNUT53zqHDbnrbe9L0Zaw91rmMBqtrO4nVD2L37kJKl9zO1wktOAX/GDusBOA3NOrDXsqtoz9XcHCGrOV6IyAsl9ygkM7etuDV7wqm1QSoDZkIBK4rqyrenczFwPvDCFGh2J8osaNQvA0Za1Gr60q+88gbPLelBymhP2tXJU7SeYfdiSw/WbkFTYjStVZt6uK//rJlnxrM3Vya5NrHyWx6Ael60SGhmwQXnITZOGtM56wLmTSE/QR8GgLP4DCRtcnmLZACu+AUaLsujisZtRWQhKBPLwkdqL54d7IXA08mI0SFYl8Zldj5938f/Qk//emxLkWhUIyGsTrcrtrUxa0PrWfdrsTob+pDcAvjJfqzFjfct4ZbL1nKyfMauGPFZlI5m9qgSU/aquqcJc+clx8udFmNn2dOfcir6UlmbQxNI2BqhbZ7GxeJdCWuK7Ect6oUlgCaov6S0BmpjmlPVLIFKDKW8ykUipEZldi57bbyrzs7venmtbXe1319EApBc7MSOwrFRDLWotZVm7q44Xevsis+MK13LB434M2uSuYcMlaaLz+wlr8/aQ7r2xLUBk2kpCqxYyc66V/9e/K7N4+rZ44A6iN+tvVmgIJbtPDSWY0RP2G/TtgPcculI5lDuq7nEl8Yo2E7suJ747WZC0I+HV3zUldjLd5VxcAKxf5jVGJn69aBz++5B370I7jzTjjsMO+xN96Aj38crrpqIpaoUCiKjKWotSiQ4hlPhBiFm/m+Tm9wXNjcmeKWh9ZjO5K+dB5/hSnpIyGlJNe+kVz7Rm/9mkFgzpFEjhq7Zw54Qifk0wr+QRIB6JoAAZm8w7aeNM1RH184dwH1dXX0Zize7Ezxvcc3kM7Zpe4qR5a/R5qAD588h7MXt/CTlVvGpXhXFQMrFPuHqmt2vvxl+O1vB4QOeJ/fdht84ANw+eXjuTyF4uBmaBFydzJXqtEpTi83NK1Uk1KpqLUokEI+g2TOQRcCQxcjOgNXi1Uo+rEL3Vyj2lc+i5Puw2yYjdkwC2lbxI57D76W+WNeR0HPIITA0DU0AbPrQ3Ql8+RsB+l6kRlNCJqjfo6cWUtdXQ1CeHUxYZ9OxG8Qz1jkbAchvVSYoWtE/SaGBh84bhZLZ9awbH7juBXvqmJghWLiqVrstLWBVSFK7Tiwe/d4LEmhUEDlIuTmWICc7bC1K12aXu6lZwban4cWtRa7fqIBwxsAiucx42pi1OJkPHDScbI7XsM//TBkPgNCYIRrqT3tw2i+wD555oCXqkJATdBAIKgL+QiYOpGAQTbvloQhQtKdzLOlM8VxdbVAef3MnIYgOUuWjvebgt2JPAtaBupnxrt4VxUDKxQTS9UJ8TPP9FJWL75Y+OWC9/lVV8FZZ4338hSKg5ORnHW3dPaTyNpkLAeBl5LShCBrOezoSdPZny1zuHVdSU8yj+O65G0Hn+4JHFe6uHL/CB3pOqQ3rKL7oX/36nPaNiJ8QYyaFrRABD0Y3WehA16q6ojpMT5+2jw0IUot3QJB0KcTDZgEfTp+XcdyJfHswF9txfqZiF9ndyIPAsI+AwTsTuRV/YxC8Tan6sjOz34GH/kInHgimIXfT7YN554LP/3peC9PoTj4GKkI2a9pOAUrGAE4UiKk1w+kaWDZkpztctXp89A0URYZ6s/Z9GUsfLo3wduyGdVIhH1lqGeOEBrZ7WsJLTxpXK8TMDW+8O7D+cgph7BuV4L/efatvbZ01wTKBZaqn1Eopi5Vi52mJvjTn2DDBnj9dS+6s2gRHHroRCxPoTj4GKkIOZt3yTsupi5wpcSn62WprICpE/Jp1AR9w9rTTUNjR0+abDWzEPaBPXnmRJa8a1yvFQsYRPwGx8+pRytEd/be0h1hXlN42LlU/YxCMTUZs6ngIYd4Qmf+fDCUNaFCMW6M5KxbFDa6BlIKmqM+zILgMTQNny7oTOXpSuX42dNbS5EhBKTytqeIJjh1NZJnjtEwk9hx7x2zZ85ItET9OJKy1N1oWro/cfrIKSlVP6NQTD2qrtlJp+HKKz1fnSOOgG3bvMc//Wn4xjfGe3kKxcFHbdBLr/Sm82TyDkXHF0Pzpom7hUiOqetltSj5QtqlL2WVIkOpvMPWrhRtfVmcCS5GdvNZ+p76OYnnfoOTiQOg+cNEj38v9Wf+07gJHQ3QhfeRc9yK9TQjTXpfNC06oumiQqGYulQdk7nhBnjlFXjqKXj3uwceP+ssuPlm+OIXx3F1CsVBxqpNXfzoqU3EMxaW46JrEDANmqKeEZ5P10jnHUI+nYA58LfKYMfdupCJ5UjyjsvO3syIHVdaIdAzHhJISolr53Hzaa82R9MJzDmqKs8cXTDi6ApRWK8rKYkaV0pm1Ab50vmLKoqXPaWk5H4qzlYoFJODqsXO/ffDr38NJ5/s/XVZZPFi2Lx5HFemUBxkDK6zaYr66OzP4biSTN5mR49DcyyAoQt0TaBrGlnbrei4Gw2YmLqgPT6y0IGCcBiHzJZrZXHTcaRjE1lyFsk1jxI5+jz8rQtGfQ5NeM7Frl3ZudjQwEUQMgWtsSBpy8FyXG65ZClHzaod+byFGp6i4Fm3K8ER02MIVYKjUBxUVC12Oju9sRBDSaVQv0AUijFi2y7feuQNelJ5miJ+gj4dn6HT2Z8lZ7vYrqQrmeOEQ+pZfmgjKzd2jdgx5LqS5pi/NBpiT2MhKmmh0fwzdtJx+l9+GH/rQozaFgC0YAyjbjr10w8d5u68N1wJeVtWvLbXeea5GjfHggR9OvGszaJpMZbO2HNtzUgDUz9x+jyOaNr3dneFQvH2oGqxc8IJ8Mc/wrXXel8Xf6f913/BKaeM59IUioODVZu6+NYjb7BmRx8IQaY3jd/QaIoGOKQhTNZySeVtLMflX849jKNm1fLRZXP5/att7OxLM6M2xHmLW3jotd388MmNzKgNsXRGDS9v9+pmqg3c7Ol46TpkNj1Pcu3juLkMVsdWas+4AiPaiND3vVNh8LUHizSfrlEb8pG1HHpTeWqCxl59b/Y0MPWm+9fwlXfP4/QaVYisUBwMVP3b6dZbvVqd117z/HV+8ANYtw6efRZWrJiIJSoUU5fiDbkn5Y13MDQAQcby6m1m1AWJ+A38hkZHMkdfxhoWrcjZDv/y21dKBchCMKyTazywureTWP0H7J4dpaGdxb929kXoCAGmpmG75ZPHj5wZY9G0GC+82cv2njQd/V6kytQ1ZtYF2NKVwnJlxfbwvQ1M3Z3I8tvVOzh18Wz0KuZ5KRSKtydV/4ZatgyeeQa+8x2v7fzRR+HYYz2xs3TpRCxRoZiaDL4hN0X8ZCwH8ByRTQ0sV9LZnyXsC5eM8Lb3pPnpX7aUohWJjEUia5edV0rG1U/H88x5gsym54d45pxIZMmZaL7gPp1fSq+tXtcEwgWQRAIm/3rRUlI5m2e39BDyGYWibJ2s7fDG7iQ3P7iOiN8g7NOZ3xwpM/4bzcDUHb1p1u1KcOQean4UCsXUYEx/ji1dCnffPd5LUSgOLoo35NqQWfDPEYW6ksIwSw1ytksm7xDP2hzWEuG3q7eX6nr8hkZnMjdh65NSktn2queZk45T8sypn0ns+PH1zHEluI4kaGo0x4Kk81666s5ntpLK2cysCyKEIJmz6erPe6MuJNiOS8jvY31bPzfet6bUVj6SV1ERv65h2+UDUxUKxdSlarHzznfC3/+9N+FcpbsVirHTk86TyjvEMxZ5x8V1vREQrisLM688EdCZzBPyeamXrV0pADKWgyZExQLj8SK55glSrz1VSllpvhCRI88hOP94RIUxDGOhWJdTbC3XNYGpe8NMe9NWWXRGSklHIovjumhCgAZ52wUpaI35aU/kuGPFZk6e10B9yIepiz2OjIgMGZiqUCimLlUn9pcuhZtugtZWeP/7vVb0vPrjSKGomu09aZI5m2xBuJiGd6MHPJ8cWyKlJBbQyVouO/oySLx2ak0I70Y/AUgpcXMpL3IjNITQCRxyFA3nf4bQwpPGTejAgNABb19Zy6ErmWd+c4TasFkWnelJ50nnHRzppfhsR2K5kt50nmTOJmhqbO5IltrL5zdH6E1bwzx1ip5EM+tCJddlhUIxtala7Pz7v8POnfDAAxCNekNBW1vhn/5JFSgrFKPFdSUPr21DE3h1vnjTuYUQpQ5HiRfZ2dGXpS9jkbW8Al7LkViOOy5mgGVryqVxrSxOohM3m0KP1BM55jxqz7iC2lP+Dj04McJAFj8kOC74DY2rl8+nMewvRWeSOZuORK6058GdWt2pPNt7MrTFM3Qmczy9qbNsinl7IkfGcnBdScZyaE/kiPh1PnDcTDXzSqE4SBhTy4amwTnnwM9/Drt3w09+As8/D+8a3/l+CsWUxHUlD7y8i9d2JagNmmiA5bjkbYe87Y4qNTWe6SsnHadv1a/pfuSH2L3tSMdG84cwapoIH7qsKnPAatA1L3VVREqv0+rady1k2YLGUnSmJ5WnI5H16nRGQCIRQuC4kv9+9i1Wbera48iIr128ZzNChUIxtdgnY4z2drj3XvjFL+DVVz0PHoVCMTLFtvHXdiXozVhoQiDlyOJlT4aA+4p0HTKbXyD56mO4+Qwgybz5N2qOfQ9awECkJu7a4IkbQwPb9UROwNBYOrOGy06cDQwM9Lz+N6/Qm86XhJE7aMRFUSu5EoQrCZo6edst1e6MNDJCCIjH4xO4O4VCMZmoWuwkEvB//wf33OPNx5o3Dy67zBM9CybmD0CF4m2H68phN9jntnSXTO5CPp3eNHsdzjlRYsPq3kFi9YPYPTtLBch6sAb/tIUIY9/NAUeDW0hb6ZrAb+jUhU0+ecaCYQM9/+GUOXzvsQ3ICu9VmQmhEDTHAuiaYHNHkjU742hClL4Hpy1oLJ1bzcZSKA4uqv6t1tICdXVw6aVwyy0qmqNQDKXSiIJ5TWHiGatkcgewsy+z39fm5tIk1zxOZvMLA545muF55iw9q+CZM7FCoChlJIAQxAIGi6fXlPnkDObUBU389C9bSeZspCsRcvj8LAE0R/1E/AauK+nMO3ypYNY4eEzESNdQKBRTm6pqdqT0HJM3bYLvf3/fhc7KlSu58MILmT59OkII7r///rLnvWLN4R/f/va3S8ecccYZw57/0Ic+tG8LUyjGSNEReX1bgrDfoDnqJ+w3WLMjzrpdCaSUZC2X7lR+QtvGhyKlJPvWq3Q//O+kNz5XEDoCs34W9WddRey4C/fZHLCILmDpjNiwOVcCqA0ZzKgN0hLzURs0+dIFi/j5R0/i7o+eOKIIiWfypPI2edtFF2Aa5WfWgJBPpz7itZH3FUTljt502feg6MWzalMXrivZtDvJig2drNkRx92f3wyFQrHfqSqyIyV86lOe187Chft+8VQqxVFH/f/t3Xl8XFXd+PHPvXfWTJLJ2iw0Ld2AbhTKVgvSslNZLIisSuFBZBGUxQ2UH4gKyqO4wAP4iAIKCD5CK4oFi0ALVgqlFLoB3aBb0nRJZjKZ/d7z++NmJplksjZJk+n3/XqFdu7cuXPOTTr5cs73fM80rrzySr7whS90eL62tjbj8cKFC7nqqqs6nHv11Vdz9913px97vf3zoS1ET6SmrPaEYvzyX+tpiiao8nvTlXuTCUW0JfF4T3OCxkgCa2BWjXfKDO0l8PZzqKRdJ0J3efFNPY288cf261JygMtnHswdZ03iy79bxvvbAziASNLCUopQ1KQ5ZgJwWGUBV3zm4C5XRFmW4jdLNuEydCxLYSowsAOc1C1UGowodKOhYSmL+qYougYHFdnfg2jCImlZ+D0OGiMJ7l24Dr/XSXNTkO3NGg5Dl1EfIXJcr4IdXbeDnD17+ifYmTNnDnPmzOn0+crKyozHf/3rXznppJMYO3ZsxvG8vLwO5woxGNpOWUUSJsFIArfDoDluku92EIol2d4QIdkmurGsgZ4oaqWUhRUOohJRvGOPJrL+LdyjplJw5Of6fSm5Q7cDmDvOmoSua3ztpPHc/OeV7GqKoQGGodk7mFv2NFR9U4y3Nu3pMsBIVZmuKPSQbNk+I5a0MjK3U8v2IwmT3aEYloLKQjfhhJU+Xyl7Dy5N01izI4jf42BiqYMRBU5ipupQgVkIkVt6nbNz333wrW/Bww/DlCkD0aTsdu7cyYsvvsgTWfapeOqpp3jyySepqKhgzpw53HnnnRQUFHR6rVgsRizWWmY/GAwC9lD/QCQupq6by0mRB1ofLUvxp7e38OBr64klLMoKXDgNaIrEiZtJdjQ0U+pz0RBJYFkWzpZVR203uhxIsZ0bMfLLUPEIKAtN08mfNBtvzWRcI8akepT1tRqqJYDouXyPgzKfi9vmTETT7Ht13JgSKgpcNDTH7Ptm2QFHnsugLN9FKGby8OINHDempNPRnT3NMZKmhdvQ8Dp18l156ZGaeNIiEEkQSZjsbY7hcxmMLPKwbW8Ep66xoyGMqRQO3a5dZKHShRjdDnA5dAwFXl3HU+hqqcDcdXuGkwPt32Sukj52/9qe6HWw86UvQTgM06aBywXtZ4z27u3tFXvmiSeeoKCggPPPPz/j+GWXXcaYMWOorKxk9erV3Hbbbbz//vssWrSo02vde++9/OAHP+hwPBAIDFiwEwqFADpsSpgrhlsfLUuxaVczgWgCv8fJ2HJft7/gUn38YGsj/7diG2u2B3FaCo8BWtyixGOg5WugFEmlwIxR6gJcdj0ZDTAH+PMqEQ6y852FRD5ZTfG4Ixgx9bPozjx0jw90A8ryu72GBoxo+Xfd0+ZWFRqcN72KiaWO9JLuDTtDGMkoR1XZCdmmpVpWXumARsyrCAWDvLdhB+MrsrcrnzgH+RReI4HL0TLd5gJ7IksnWqARiia5+JgaDqssxEJxz4trCcWiVHoVDj29MTuWgrg9g4ZHT+JFI6KS6LqO26mTX6R3257hZLj9m+wL6WNu2Jc+pgYrutPrYOeXv+ztK/rH73//ey677DI8Hk/G8auvvjr99ylTpjBhwgSOPvpoVqxYwfTp07Ne67bbbuOWW25JPw4Gg9TU1OD3+yks7P8qsakAyu/35/QPKwyPPi7dsJtHlmxkU9vVUiPyufZEO2cj27JxXbf3Znp/ayN3v7yJhnCCUMyuBaOh2fs1NdlDF/Fk5yFCar+r/qYsk/CGtwmtesUezQFqP15JbPRncJUUQi8WfqW+e9t7UWenPhJn+xvbWPhxIH0fQzvjbG/WGFHgtAPJlqmn5oT9GsvSqQ9bhHDh72SjvSMLCskv3MaHtU1UFjozfraUUtQFExxW5eeSEw5D1zUsS6G5P2XzrkYchoahta7BMC2It0SbOhamGWN7WAPNDsBKfW7CCa3L9gwnw+nfZF9JH3PDvvSxp+f3OtiZN6+3r9h3b7zxBh999BHPPvtst+dOnz4dp9PJ+vXrOw123G43bre7w/HUaq6B0Ha1WK4aDn1cumE331uwmlAsSXGeC5ehEzct1tWG+N6C1Vx23CiWrN+dsWw8lbx63JgSnluxnaaYSaHXTVMskt7iwTAUCVOhLNUyCZTdQIzsJPZsI7j8BRJ7t5EKTwxvIflHfA5HcXWX7elMeguHTl6bOqq1bN6pFBiGkb6P95w3lVKfG4ehEzMVHr3jws+oaeHQ7SCjs58Zw9C4btZ4bp+/itpgnKI8J25DJ2ZaNIYT5LsdXDdrPEbL/lmGoXH65Eo+2BYgaQIGLdNq9jRiqj9WS98MXcNCI5xQRANRCj2OLtsz3AyHf5P7SvqYG/rax56e36ftIjZutDcDveQSqK+3j730EqxZ05erde93v/sdRx11FNOmTev23DVr1pBIJKiqqhqYxohhy7IUDy/emK5143Ea6LqGx2lQWeimIRzn54s+Zu2OQNYly396ewvbGsIU57lwtGzYmbRUehuD1KiNY5DyPax4hOC7f2PvK4+Q2LsVUGiaTt6EGZTO+Qbeg6f1+4ejhh3cuBw6esufhqahWo5XFrpbcnE2MrGyoNvNOMeNyO92M86utn3IllB8wvhyivJcuBwGlrI3DLWUwuPSM7an0HX7g1LXNLuSs2mv9ppY2Xm+nxBieOr1yM7ixTBnDhx/PCxZAj/+MYwYYW8X8eij8Je/9PxaoVCIDRs2pB9v3ryZlStXUlJSwqhRdsn4YDDI//3f//Hzn/+8w+s3btzIU089xec+9znKyspYu3Ytt956K0ceeSTHH398b7smclxqZU9xnqtjEKBBPGknsFb7PSgFzfEkDl2nosDFzqY4z7yzBUxFXFnsaoqnVxUlLYWuga61BkADLbJlNaEVf8eMBEntHe4srqbwmM/jLK0ZsPdVLf+JJy0chj2FZ6HQNHDoOpqmUZTnZGN9iHV1TVw3axy3z19FXTCWZVTG4LpZ43qUDNzZtg/ZXju5upBJ1YWsqw3i93hakpR1FIrNu5tb+6LsPbXaVnI2NI11dU1MHTn8p7GEEK16Hex897vwox/BLbfYu56nnHSSXXCwN5YvX85JJ52UfpzKo5k3bx6PP/44AM888wxKKS655JIOr3e5XPzrX//iV7/6FaFQiJqaGs466yzuvPNODKN/a4eI4W9v2K6m6zKyTKnELRKmnb26IxC1A5mW5cpuh06h10l9MEaJ02J7IIJCw6FrJFsCHkvR5UaV/clKxEju3YYZsROBdacH39RTyZswo99r5rRnaPaIiGnaIyFg1w/yOnU8Tvu+ug2dgKXYG44z65By7jlvanp5fsBSOHWNiVUFva5ro+taj4KQ1J5at89fRSCaTAdZDeF4y35cGk5DQ2GP+qCBx2lQmu8iHDfZG4739fYIIYaoXgc7q1bZ+2K1V15u19/pjdmzZ3e7+umrX/0qX/3qV7M+V1NTw+LFi3v3puKAVZLnwmloxE0LT7ugIGlZWJadyxFPWjgN3c71ACIJi3gyRp7TwHQoTEu1VPHVMGitGzPQlGVihYNY8TDukZOJbV+Ho6jSrpmTNzgjEaYCreXfrIKW4FGjvMCTHi2LmRZOXaMkz65o3JtRmf6SmvpqG2TRsqt6eYGLEp+TfC2B5tIwdAOPSyeasHDqVrrdQojc0etgp6gIamthzJjM4++9Bwcd1E+tEmIATK4uZNyIfNbVNlFZqGdMZelaa0Vep66lp6Q0wKlDwrQIJ5IYHg1d14glUym8Ay9Wt4HErk9wj5wMym6l4Sui+LRrMfppi4dsDF0jVb+PliRfsKd8UjX9NKAs396TClpzcSZWFWTk4vR0VKY/tQ+yirxO/vvlD/mwLgQK3E6DAo+BQuu03UKI3NDrBOVLL4XvfAfq6uwhfsuCf/8bvvlNuPzygWiiEP0jNb2R7zaoC8aIJEwsS7UUpbPXQ9u/xDUspUiP17QZgDCh9bf+ADMjQQJLn6Xx9ccIrX6VZGMdmtONw1+B4S0Y0EDHZdj5K9loGhT7nDi01lVbqftYF4z1KhdnoKWCrFmHlDOtpojrZ49Pf/9jSXPItlsI0b96Hez8+McwapQ9ihMKwaRJcOKJMHOmvUJLiKEs28qe5mgCj8tIj1bETYtY0v6KmxYJU6FrGm6HQaKlho7boeEy7Lwdp25PZ/UXZZk0f/wf9vzjV0Q+XYlqGc1J7N6Co6AMzej1gGyXdA08Tp2DijwZC801LXsVZQ0o9Dip8HtxO3QSptXtCqmhIvX9P6yqgGjCHDbtFkLsm15/ajqd8NRT8MMfwooV9sjOkUf2z15ZQgyE9kUCZ4wtZcbYUlZtD/D8im387YMd6ZGdtpQCM7Ws3CBj2bLZsky5v8d4Enu3E3xnAYm920nXzPEUkH/kHDyjj+jnd7MpoMrvwed2sDsUJ94y4mHXvGgpgtiuo4amEUqaHD26mPOnj6Q2EOGgojzOObwKh6NPFS0GzczxZRw3poT3NuwghItSn3vAc4iEEPtXn/8XcexY+8s07aTlhgYoLu7Ppgmx79pu1Nm2SOAJ48v48/KtbNrd3KNZqYQJTWaS0ny7pkx/Fwe04lFCqxYRWb8MpexVYZqm4x13LPmHn4buzuvfN2xD1+xKwxoalX43W/aESbasLtPabLiZooBdoRh5LoNgNMHP//lR+t4+/962YbF7uK5rjK/Iz+mqtEKIVr0Odm66CaZOhauusgOdWbNg6VLIy4O//x1mz+7/RgrRF0s37Ob2+as6VEte8ele3li/u9fXU0BcqX7f7iER3E3jq79tqZkDoOEorsJ/zNx+rZmTJW7BYdhbLDRG4iQtB5YFhR4HTsMkkVRYWSaylIJowsTQNWoD0XaVqGX3cCHE0NPr8ea//MXeBBTgb3+DTZvgww/tIOh73+vn1gnRR51VS06YFuG41f0FBokVj6DMOJrT3vNNc3oomH4Wpadf3+/FAbPFaHZ1YdgdivPJnjDbGiLousZlx9aQ73agYU/fOXQNb0ul6TFleSggFEtSUeDuUIk6VUHZGoTiikII0RO9HtnZvRsqK+2//+MfcOGFcMgh9kjPr3/d380Tom+yVUtWKHYGY4O0YLxzSllgmZjhACoRQwN8U04mtnUNBdPPGrSaOVnbBiSSJn9fVYfLoTEyz4uhazh0HY/LnuqKxM2W+lj2Enxvm7I0bSsor9kRlErEQoghodfBTkUFrF0LVVX2flgPPWQfD4dBihaLoSJbteRo3CLeUiV5f4nVrqfp3RfwTZqNo8j+vwbdk4+nuArv6O73fhsoGvbqq9SjaNwkGLUYUeBNb7KZkrSsdHXppGVBu7VobSsoCyHEUNDrYOfKK+3RnKoq+8PutNPs48uWwWGH9XfzhOibbNWSk5Y1WHUAOzAjQZpWvEh0yypA0bz2dfwnXIYjvxTN4eyX9zD01H5P9p+pHB2HbifkKmVP77VNrjZ07ORkjYyl9y6HA8syaYolKWpXUdjeA6v17+21r6AshBD7W6+DnbvugilTYOtW+OIXwe22jxuGvW+WEIOt/dLyydWFGdWSKwo1Ygl7k0/Inqg7UJRlEl6/jOZVi7AS0ZajGrq3EMNb0G+BjsepM67cRzRusSsUp6LARX1TFEPXyXM58Oc50DSNPaE49cFoej8vQ9MwMpZcKyyL9K7mgUgSv9eZsWLJ7dTSj92OzARmqUQshBiK+rT0/IILOh6bN29fmyJE73W2tPy6WeO4btY4bv7zSj7eGUIp1VI3Z/DaFt+zjaZ35pNo2JE+prvzKThyDp6Dj+y3Jc8aUFHoIZZUBKJJvC6dhKVoiplo2KMzjRGD8gI3PpfDHuVBYZktAU+ba6Wmp1B24UCnw8i6Y3mpz4UCdjbF92k3cyGEGAx9Cnb+9S/4xS9g3Tr7g/Gww+zVWKee2s+tE6ILnS0tTy1/vuy4UQDpQEdhLz8c6LVYVjxK6P2XCG98J72XlabpeMYeTcG0M/q9Zo7D0AjHTJyGRZXfTX1TjMZwHF3T0HXQ0IgmTLY3RKgu8uB26IRjdu6SfW/sAoIKe5NTj0MnmrSYWO3nmhPH8Zslm7LuWA70y27mQggx0Hod7Dz4INx8sz26841v2Mfeegs+9zm4/3644Yb+bqIQHbVfWp4aJfHoBpWFOnXBKP/z+kZM08LtMIglrcxElgGizCSN/36aeN369DFHURWlx52HXlKTfuv+aIZT1xhV4iUQTXLDSeM5YlQR//3yR9QGohxU5OXTvWEiCQunbgdESVOxOxSnNN9FOB7B0OyCggnLQte0dNVkh6GR5zK49sRxHD+hnJnjyjrdsXywdzMXQoi+6HWwc++99qhO26Dm61+H44+3982SYEcMlLa5OXtD8Q5Ly1M0za4Js70xgq7ZG3s6Wn4Bx5IDE+kopbCiTViRJrxjjiKxcyMYTvKnnkLeITMxDAcOvWVpt6m6DXS6C4Z0DUaW5OFxGgRjJgeX+9A1jU277Hui6zrlBR62N0RIWAqHrqFrEE0kCUQ0ygvcjChws6MxQjCaxLIUuq5R6HEyqbqArxxbmR6d6WrH8v2xm7kQQvRWr4OdYBDOPLPj8dNPt3dDF2IgtM/NMS2LplgSp0PH4+xY8yC9pxP2Dt6aprU87t9gR5kJkoHddvRh2VNDDv8ICo7+PO7qQ9M1c5yG1hKEwM5AtEe5QxUFbhoiiXRidYrboVFdlEe+20EkYaZXPrVfbp/vdnBQsZddTVFiSXu5uKXgoCIv3/vcxPSozO7mGI3NCYrznJTmu5lUVUBTUzBbk4QQYljqdbBz7rkwfz5861uZx//6VzjnnP5qlhCtsuXmBKMJGiMJtjdE0Es08t2ZP8rRhB0g6BqtRQX7eVAnVvsxweV/BcD/mYvQdAPdnYfuLcRZXJVxrmkpdodiaBoYuoZpKrt8eZugrOWh3W5do8DjpKzAxYb6ZmJJi1KfkyKvC6/bQEPrsPJpzY5gh+X2+W4HPpePaMKiOZ4kYVrcc95UptUUAWQdlVH9faOEEGI/61Gw07Yy8sSJ9nTV66/DZz5jH3vrLfj3v+HWWwegheKA1llujt/rZG9znHDcpD4YxVfuQyMV1ChiCRMNsCxQhrKDg34a1UmGW2rmbF1NaqQotmUVvqmnojvdWV+TCnKUIj1SU5zvpMDtxNA1djTaoy9Ow57AMi27LpBbGTgNDUvZOTdods5z1DQ7rHxqu9y+slBP3ytN0/A4dRojiolVhUw9SKadeiNbaQPJSxJieOlRsPOLX2Q+Li62qyivXdt6rKgIfv97+P73+7F14oCXbdsHsH+Bjyj0sK0hTCRhEggnKPQ408uffW4HCgjHTRKmgpaVRvtCKYs9a/7DnndewUrG0sdd5QfjGXd0OtDpKt8mtdeUqaA5alLl97bsNm7n15iWShf4C8WSbG+MYFkKp0MnnDDZujeM1+XA5zI6rHzSdY3rZo3j9vmrsi4XlyXhvddVaQNZcSbE8NGjYGfz5oFuhhDZZdv2ISXf7aDa72VHIEI4bhJNWunlz9ecOJafvvQha2uDmP2w1jy+ZytNby/AEapDtYzM6G4fBUfMwTNmekYgli3QSZiKpKnQW0Z4DOzppkjMJM/tSOfX1AejRBJmugAg2IFPPGHZQZBmJxpff9J4Lj12VIfAZeb4Mu45b6osCe8H3ZU2kJ3dhRg++lRnB+wNQTUNSkv7szlCZMq27UNbTodOmc/Ft8+cSEm+Kz3N8NamPdQ3xVCqJXe4j4M6VjxKaOVCwhuXo2HhcOiAhnfs0RQccWavauakqhaX+9y4HTpbG8LsDsUo1zXcho6ha3hdBoVeJ5ZS1LdsWqraXqAlT+exf2/moGIvZT53h2mVmePLZEn4Puq+tEGMhxdvZMbYUrmvQgwDvQp2Ghvhe9+DZ5+Fhgb7WHExXHwx/OhH9lSWEH3RWV5EZ3kokLk1weePqE7/0kn9ojItxaiSPOqC0XTCckpP6tyoZIL4rk1ENi0nVYrQU1RB3hHn4Cwf06d+2tNTCXxuD8V5TmpKfNQHo21GYAo5fVIFd/99bUZNnlSDU8HPxl3N3PzMSvJcRtZpFVkSvm86mz4F2dldiOGox8HO3r12QvL27XDZZXaislJ2FeXHH7erKi9dagc/QvRGd3kRvc1DafuLyuM0MDSNT/c2g4JUmR17g0wNs6W6cltKWVjhIFasGcPrx3PwkUS3riJ/8smMnX48O6JGn1OdLSASt9gZjHFEjZ/HrziWdXVNGUHeE//5pCXPqO1O5B25nTo+t0OmVQZAV9OnIDu7CzHc9DjYuftucLlg40aoqOj43Omn23+2T2YWIpvUSM6bG3bxh/98SjxpUuJzd5oX0Zs8lPa/qLwuA4/TIBw3M7aLSLab21JmgvBHS3FWjEXTWn/J+SbNwjf1VJw+P1rHmbReM5UiHE8SiCR4+5O9Hdpf1xjt+CLVcSTKtBQep0yrDITupk9lZ3chhpceBzsLFsBvftMx0AGorIT77oNrr5VgR3Sv7UjOrlAM01J4nQZJS+FxalnzInqTh1LkdaJQNITj5LkceJxtqgl3kq0c2/4RwXdfwGzeS974GXjHHY3mcGH4itCM1M7k/Vd/ptTnpjYQyzoiU1HkSf89vTFnFqmq0DKt0v96On0qO7sLMTxkH6PNorYWJk/u/PkpU6Curj+aJHJZaoXLutqgXXfGslcoRRIm2/aGaYomgI6/wKE1D2XWIeVMHenPGui8uX4Xd/51DYFwgh2BKJ/sCbF5dzPReBKfu+OPezIcoPGNJ2lY8jhm814AIp+sQHN5cBSWtwl0+o/XqVNR6Kay0E0oZvLw4o1YbUaZjhpVnA5kgPQmpu352hRSdBv2TucyrdI/Usv48932ru+RhIllKSIJk7pgTJbxCzHM9Hhkp6wMPvkERo7M/vzmzbIyS3St/QqX3aE4LSVwWlYqKbbuDVNTkkeBx5mRF9E2gbnIawcgjZFExgjPb5ds5OeLPs7YXiFpQTJu0hw3M9qiYxH+8N8EV/0LzHg6YdlVNpqCY+biyB+YH2YNqGizuifbiMzUg/xMrCpg9fZgp2NJXqeO19U6vSLTKv1PlvELkTt6HOyceaa9EmvRIjt3p61YDO64I/ueWUKktE0cbo6b7Gm2l1ZrtK6OMhVsawgzoqBlKkfBlj3NzHtjExvrQzTHTcLxJEqB26FT4DaoLMrj0Ip85q/cQSJp4XRo6GgkTCvrHlTx3Z8SfOevJBtr7ffWNHR3Hr5pZ+Ade0yH1Tf9qdjnpMDTOlqULdFV1zVumzORm/+8kj0he/m8Uq25RoYGlS3FCEGmVQaSLOMXIjf0ONj5wQ/g6KNhwgT42tfgsMPs42vXwkMP2QHPH/84UM0UuSCVOOw0NGoD0YwaOFqbteBJC3YE7CRdh67x43+sw+3Qceg6gUgifb1k3CQcN6lrirNyayOQ2ltK67DnFNhBQdO7LxBevyz9Zgpwj55G4fSz0d2+Aey9LZawCMWS6b28OhuRmTm+jF9ceAQPvb6BD+ua7J3SlSJuWrgddk0ey1JSHXkQyDJ+IYa/Hgc7I0fCf/4D118Pt93WuqmipsFpp8GDD0JNzUA1U+SC1AqXpmiSWNLE0FtGYFp25G7PaKk2HEtYJJJWOnhpWyOn/csUkEhaOAwt4zkrHsEMB1BmMv0qR2E5hcfMxTVibMY19qUIYbrtutZhewpDg1jSYntDhIOKvfhcRpcjMtlGFQKROL9ZskmmVYQQohd6VVRwzBhYuNAuKLh+vX1s/HgoKRmIpolck1rh8sHWRixLYTjsjTE7q/DncztojttBUWqZeE/GLRSty8qVmcQMN6IS9l5W3rFHk9j1CZ4xR+KbOAsty7LingY6nRUmdBqk6+SkztNa/qPrYJoWdYEoPrdBvtvR5YhMtlGFmePKZFpFCCF6oU/bRRQXw7HH9ndTxHBnWYpV2wI0tEscTkmtcLn1/96nKZYkaVoku9i3qimazDpy092vdQ0wE3GaV/8LlYjjHXd0axu8BZTM+Tp6llVWPamq3FlbUq8bWeQhmrTY0xxHB1wOg4pCN5qmsavJ3tkcTSOeNDm0soBvn3For0dkZFpFCCF6p8dLzwfCkiVLOOecc6iurkbTNBYsWJDx/BVXXIGmaRlfM2bMyDgnFotx4403UlZWhs/n49xzz2Xbtm2D2AsB9pLyO19Yw3VPLuebf36fa/64nHmPvc3SDbszzps5voz/vuBwCj3OLgMd6Dzw6C4giWxbx54Xf0nzuiVENr6D2dwAuoGRX4ojvyRroNPZdVO7lDs6+ZfSdu+qfLfBq7fM5oaTJ1DodjK61Me4ch8FHif5bgcHl/oYXeJjZJGXAq+Tb5w6QaaehBBiEOzXYKe5uZlp06bx4IMPdnrOmWeeSW1tbfrrH//4R8bzN910E/Pnz+eZZ57hzTffJBQKcfbZZ2OaZidXFP1t6YbdfH/BKjbvbsbndjCiwJ2xjUH7gOeECeXMPbK639uRbG6kYckfaXzjD5jhls3bUJjNjTj8FeguT5evz8bp0HAZOk7DoP3OAU5dw6HbAZFD1/jGKRNwuQyOGlWMz21g6FrGyi5Nszf6dDp08pwGZT533zsrhBCix/q863l/mDNnDnPmzOnyHLfbTWVlZdbnAoEAv/vd7/jjH//IqaeeCsCTTz5JTU0Nr7zyCmeccUa/t1lkals7Z3SZi70JA0X2KsipKa2lG3azcHVdhymgvrfBJPzhmzSveQ2VjKWPO0tGUnjMXJwlB/X52smkwumwJ7j0liVjqRwcU9l/9+c5+drscVx94jhAqu8KIcRQs1+DnZ54/fXXGTFiBEVFRcyaNYsf//jHjBgxAoB3332XRCLB6aefnj6/urqaKVOmsHTp0k6DnVgsRizW+ksxGLQr9CplL+/tb6nrDsS197fV2wNsqm+iJM+eGmq7BkrToDjPwab6JlZvDzB1pL8lONpAPGHi0BWaZicf9/XWxOo/oWn5CyQDdS3vD5rTS/7UU8k7ZIY9/Ynqc0ClgKRpggJNUxR6HNx9wREEwibbG8McVJTH2YdX4XDo6e+vpsG1J47l+wtWsTMY7bB5aYHb4NoTx9oBk2kNuWTjXP55TZE+5gbpY27Ylz729DVDOtiZM2cOX/ziFxk9ejSbN2/mjjvu4OSTT+bdd9/F7XZTV1eHy+WiuN1W6xUVFdR1sXfFvffeyw9+8IMOxwOBwIAFO6FQCGBAC9btD7v2NlDiNCn16hQ5FK1jHwCKKBZGMskHm3Yw0qfYtKuZUDDIYaUOdjUliSYzV1n1+O5bJoFPVtG45C84AEdLUo1/zFQqjz0Lhze//zqJnbdT6HFw86xRHFmZqoBsv0dzc1OH8yeXO7nrzLH85d1tbGsIk0wq8nWNw0bmccFRI5lc7mTJ6k9bnzcVDkNjZLH9/LSaon5tf2/k8s9rivQxN0gfc8O+9DE1WNGdIR3sXHTRRem/T5kyhaOPPprRo0fz4osvcv7553f6OqVUlzfstttu45Zbbkk/DgaD1NTU4Pf7KSzs/6mFVADl9/tz7oe1vAT2JgziEQ00jYaWaaxQLMmuphjRRBJLwS+XbOW51XsZV+5jY8Ci2GuwsxliSfs6XQU7mkbGyI8VbcaMBFFGKXGXHyvcgFFQRuFR56BVTWCnBTT3bz8dOiR0A8Od1+Pv44l+PydMGpV15Gbpht3c9dImQrEkxXkuXE6dZtPirW0RVu/axI/mTt1vycu5/POaIn3MDdLH3LAvfezp+UM62GmvqqqK0aNHs76lyE9lZSXxeJyGhoaM0Z36+npmzpzZ6XXcbjdud8fk0NSKr4HQdkVZLkjvVdUcp7zQy9Y9IYqdDoJRRcxU7ArGsFqK6DgNnea4xcqtAd7bGgAgGLUTyNNbHtDF6qtUZeVwABIxMBP2ubqBb9IsrEiQvMmz0Q1nP+5LbgdZDk3DAqqLPITjJs+t2M5nJ4/OOtXUdv+uVGBjGDqHtxulsSzFI0s20RQzqSz0pn8m3LpORaG98eQjSzbxmXFl+21KK9d+XrORPuYG6WNu6GsfczLY2bNnD1u3bqWqqgqAo446CqfTyaJFi7jwwgsBqK2tZfXq1dx33337s6k5bemG3enNEROmIpY0aY4l2K4lqI1omJYduOga6C05OZbKnjeTqlWT2uXbUgqnQ0dTikjLFJeViNK86lXCG5bhn3EBRr5dxVIznHjHHInm6P/NLzXAaeh2KKYUTsOgKE9nW0OYNTuCHQKYtvckkjDRNY2akjy+efohnDChPOPctnuEtf+H2n63d6mnI4QQ+26/BjuhUIgNGzakH2/evJmVK1dSUlJCSUkJd911F1/4wheoqqrik08+4fbbb6esrIzzzjsPsIe8rrrqKm699VZKS0spKSnhm9/8JlOnTk2vzhL9a+mG3dw+f1V6+iVuWjSE4+nRF2WpdM6OpewgJmlZXSbjpCoeaxo4dJ0qvxddg11NMQKbP2DP23/DDDcC0Lx2MQXHzMWZ58eZl09S9f//6WiA06GjAwlL4XXqeJw6KEUymblpZ9t70hCOE08qEqaJUtAQjvOVPyzn1tMOSa/UgtY9wlzt17K3yLY5qBBCiL7br8HO8uXLOemkk9KPU3k08+bN4+GHH2bVqlX84Q9/oLGxkaqqKk466SSeffZZCgoK0q/5xS9+gcPh4MILLyQSiXDKKafw+OOPYxgdtwEQ+6btMvPKQg9oUNuyYafLoaFrCoeho0yFYdh7QyVMC8OAVNmjVP6No80WEGAHPB6HTqXf3jNqy7YdhFf+nYYNq7DMlvBJ03GWHISzsBzD5cbtNDDjZo+nrjqrkNz+eKp2TsJSGJpGeYGdkBxNmuS327QzdU8awnHCMbsthq6jafZzsYTFzxd9zMSqwvQIT2qPsLhp4cmyXUVnm4Om3m+ord4SQoihbr8GO7Nnz+5y9dPLL7/c7TU8Hg8PPPAADzzwQH82TWTRfvolEjdbN/TUNAzdwlIKXbenr9AgoRSoNsGEsgMJ+zUQb9lDStegzOdCs5KseeNlAqteRbcS9vU0MIqqKTzqXDwVB7cEVhqjS31cdEwNz769hXV1Teld1J2GTixppd/TqWsYhoalIN6ubLOhtbRNgdOw9+pKKoVD2SM65QUe8t2OdH2cw0bmZdTHSd2TeNKepnMYWjoPydA10CziSYuf/fNjZrbk4PS1Dk/76UOnoTFuRL5sAiqEEN0YVjk7Yv9qP/2StOzdyrWW2ZjUr2yHoZM0FZrWshmnah09SeXyKOyRkxRLweaP1xBd+SKO8G4qfS627rVQhof8KbMpnHgChsOBUhBP2leqC0SZNrKIaSOLeH7FNt7YsJtgJIGGhhmx2+rQ7faAHdg4DS1jk04LcBk6I4u9nD99JKFYkufe3UY8aVGa78LjMIgkzHR9nAuOGpkxkrI3HCeSMEmYpj2i027nLh0NTYMte5rTOTipPcJun7+KumCsQx2efLfRYXPQ9tOHLkMnblrpKtX3nLf/Vm8JIcRQJ8GO6LH20y+Oluma1M7lCnuaqsTnYndTnKTZGuy0pWsaiZaRFw3AikM4QNM7CzBDeyjLd2MpcI6cRP60OXiKSkFp6aDJYdivb4zE7VyZ5tbAZmRxHmdMriDf7eBHL64lnlTouoXeUu7QUuAyNJyGRkWhl+MnlLFqayM7g1Ee//cnLcftlXp7QnGC0SROXWNiVQHXnjiWyeWZ+2qV5LnQtZa2ZZlNUth7sliQkYMzc3wZ95w3NT1SE7BU+n3aj9S0nz5MjQR1VaVaCCFEKwl2RI+1n37xuHTcDoNowkTTFaYFbodOScvIw/bGCLqyV1ilpq80jXSujlIWKhJAxSJoGngO/SyBd+azy8rDP+lzFBw0CU3TSCQVqbTnVBIzGpgWbNkTptLvSY90bGuI8Ow7W7nnvKl88/RD+fmij4knLTTNDjrsfa50SnxOLjm2hqeWbekwWlIbiOFz6Vx/0nhqSvLSuTGaZheebH9PakryaAjHsSxlT121UEqRtOyRMG/LfWlr5vgyZowt7TYHR1ZvCSHEvtmvG4GK4SU1/ZLvtmvBRBP2VA/YU0uaBqU+N9GkRShmclCRl7vOncy3zjiMaTVFlPicLftLQWLvdhI7PsJIRlDYQYGj5CAKj5lL6Rk34ho5Of2LPTUwlBqZiZsWZkvqTYnPicdpoOsabqeO3+OgoTnOfS9/xJUzx/Do5Udz+Mgi/B4HPo+DIq+DaTV+fjR3CkvW706PlqSu4XEaVBa6aY5bvLymjs+OL0tPPXV2T755+iG4HDoJU2EqC6XspfZ2grOdQzS+IvteWLquMXWkn1mHlHf6Pj1ZvZWQ1VtCCNEpGdkRaT1Z6dN++iVhKfxeJ5ayyHcr6mMmDlN1mI655sSxrNkR5KX3NnH/Q48S3bAMb/EIOGouVsv8l+EtxFV6UNZ9srJVWLaTke3VTKmKzbGkiWUpVm1r5ILf/Idvn3Eoz183s0O/+nO05IQJ5dx62iFdjiK1z8HpjX1ZvSWEEEKCHdGiNyt9sk2/HFaRz/ubagnhotTnzhIoKXavX8Grj/4P4Y83YGga0b21JLd+iGv0ETh8RWiG0Wk9ntThtsvEXYY9lRaKJdneEMFSLdNIOpiWYv3OJm57/gPuPf/wDn3o71o3V584jolVhfzsnx+zZU8zFuB12CM6+7paSnZRF0KIfSPBjujTSp/U9EuKUorxFflZ9zb59NNPeeSRR1i2bBlKKbxOg3DCovTQY7DGHY3m9bdco/M2OlqWtysUSdNe5u1z26Mcu5piLUvUNRKmRWqRVyxpsiMQ5d6F6/jr107ICL4GYrTkhAnlzBxX1u91cPqyeksIIUQrCXYOcP250seePgrQEElQkufiYL+DXz/6B/721+ex4jF8bgNN0zjm8IlsrzoBs+hgmmImVpdXbR3NMVuqLLsdOqalMBVEYnatH02DhNlaW0fHXvVlWoq1tU08/fYWvjRjdPqaAzVa0j4I7EpvCgT2ZvWWEEKITBLsHOD6K3dl6YbdPPPvj1hRGyWWsAjVfkz9W38j3rgTsPNrCvPzuWreZXz3+it4b3sz9/xjHat3BLttY0t9QrtOj7KDnJElXiwFu0MxLEthqcypLoehY+gammYRTyr+9PYWLj12VDqY2N+jJX0pENjT1VtCCCEyyWqsA1x/rPRZumE331+wis27m3GrOHrzLhr37CW8t5akZWFaFkbVoTg+exULo+N5b3szM8aWUpTnsvec6kaqMnJqBxDTUoTjJl86bhRjyvMzdkxPVVBuXQKuYehQH4yxpl1glRotmVhVQDiWpD4UIxxLMrGqYECL9KWmDdfVBvG5HYwocONzO9LThks37O70tT1ZvSWEECKTjOwc4PY1dyU1DRZsjlDhi9IcjrG7KYxRVIln5GQSDbUUTjudgoOnYqKzOxTn3oXruPvzU/iwLohltRbd64xDt7d60ACvy6As300oZrJk/W7+fPUM5vz6DTbubsZpgKHpbZasK0xL4XbY/coWsA32aIkUCBRCiMEnwc4Bbl9zV1Zta+SDD1bRvP4trONPJpY0W7ZzAN/Ez2L4SjC8BWiGjgNIJC3W1ga59c8r2dPculu6Bti1AlsXmSdbAqGKQg8OQ8Oh27uPa5qGw9DZWB/io/oQV54whjtfWINlKex4TaGUPQKkaxpFeS6UUp0GbL3Js9lXUiBQCCEGn0xjHeDaFwqMJOw6NZGESV0w1mXuyt69e3n0fx/ikxcfpmH9CurWv08kYScJ6+48nGU1GN4CFHYhwHjSwsKufPzJnnDGMnPVcjxp2QX50kGQZm8PUeBx4nUZ6QCh7fTapceOYlJVAbquYVoWSdO+hsdpUF3kIZa0GDcif0gszZYCgUIIMfgk2BG9zl2Jx+MsWrSI66+/nmWvvgzKRClF7Ucr0XQHRkEZhq8YTdNbqx+rzBI67R+3ZSlos1envT1EO22n13Rd47Y5E6ku8pLvcTKiwE1NcR6Vfnu6aygtzW47bZiiUETiJk3RBMFoQgoECiFEP5NpLAH0LHdFKcWmTZv47W9/y9tvvw3YtW7yXE6MmqkcderZ7NKKaNgTaVk51Xr9tsUAgZZf+F0U1km9TgO3IzNIyTa9NnN8Gfe2WZrdFEsOyaXZ7acNm+NmRuVnBRR6nQQiMrIjhBD9RYIdkdZV7kowGOQvf/kLzz33HJFIJH187NixXHLdBTy1ySCkK3uPKodONJmZctw+rLGnozoPdnQAzf7Fv7Mp3qOl4cNhaXbbJe9bG8I0x+xRMV2319cbmoZlwfcXrB7QFWFCCHEgkWBHdCmRSLBjxw7uvvtuNm/enD6el5fHBRdcwPnnn09hYSGT1+9K19lxOw3ibSoZt+cwtIxRH4euYbbJ0zE08DgNHIbO1Z8dy9KNe3pcSG8wk437aub4Mn40dwo3/Om9dOVnAK/ToLzAg89lyKosIYToRxLsiKyUUuzdu5e9e/eilKKmpiYd7MyYMYMrr7ySsWPHYrQUv5k5voyJpQ62NWs0RBJs3Rvmz+9sYfX2IGh2/RuXoWGhkTAVaK3RjqFrGEDSVDgdOgf5vWg6hGMmJ4wv55oTxw3p0Zq+8HtdeB06BUVeDD1zpRkgq7KEEKIfSbAjOgiFQtTX15NMJtPH5s6dy86dO5k7dy6f/exnycvL6/C61KhK6hf2xUfXcMFv/sOmXc2U57vwugya4ybbGyKYKjNB17TA0HWq/F68LSMbqZyc/h6t6c02DQNlbzhO0oISnzPre/d2I1IhhBCdk2BHpCUSCVavXs3vf/97Jk2axCmnnJJ+zufz8cMf/pDS0tIO9WEsS7F6e4BdexsoL4EpB9mVfR0OnW+fcSi3z19FIJpE0zXynAZlBS52NcVQLaM7SdMu/DeiwI2ha90ued8XfdmmYSAMxEakQgghspNgR6CUor6+nqeffpqFCxeSTCbZsGED06dPp7i4GI/HQ0VFBW63u8NrU8HDpvomSpwmexMGY0e05tR0toHlcWNKOHNKFU3RBC+v2cnOQIRwwsRpWgO2gqovu7sPlIHaiFQIIURHEuwc4Jqbm1myZAlPPPEE9fX16eN5eXkEAgEOPfRQioqKsr62bfBQkuekxGsQi2gdgofuVkn1NSenN9NRQ22bhv29EakQQhxIJNg5QCWTSdavX88TTzyRrpkD9pLwU045hYsvvphRo0bhcGT/EWkfPOgaaJqJx2lQWdhxNVFXeTd9ycnp7XTUUNymobNRr6FWG0gIIYY7CXYOMKlVVvPnz2f+/PkZNXPGjBnDvHnzmD59Ovn5+V1ep2Pw0Lq6qjfBQ1+ShfsyHdWTbRr2R0LwcKgNJIQQw50EOweQcDhMfX09//jHP/jLX/6SPp6Xl8fcuXM599xzKSsrQ2/ZnqGrQKQ/goe+JAv3dTpqKCcED4faQEIIMZxJsHMASCaT7Nq1i6amJgBOOOEEFi1aRCAQYMaMGVx66aVMmDAhIwG5u0BkX4OHviYL93U6ShKChRDiwCXBTg5TStHQ0MC6desoLy9PH/d4PFx++eW4XC6OP/74DgnIPQlEZowtbRc8ZL5vV8HDviQL93VESRKChRDiwCW7nueoSCTC8uXLufPOO/nhD3/I3r17M54/7rjjOOOMMzoEOu0DEY/TQNe1lsRjexfxhxdvBOC6WePId9vJyJGEvcdTJGF2WyenN6Mz7WXbNbytrkaUeru7uxBCiNwgIzs5xjRNduzYwXPPPcfChQtJJBIAPPPMM1x//fU4HA5GjBjRaQJybwKRtquJNtU3sVeZhONGt6uJ9iXfp7PpKIUiEjPZHYoxpjyfiZUFWa8tCcFCCHHgkWAnhzQ2NvLmm2/y5JNPZtTMKS4uZsaMGfj9fsrLy9MJyNn0NhBJBQ+tFZSL0xWUO7Mv+T7ZpqMSpkV9MEYsaQKwdW8zVz7xzrDeLFQIIUT/kWAnB0SjUdavX8/TTz/NsmXL0sc1TePkk0/mggsuYPTo0Xg8nm6v1ZdAJBU8BArA7/d3GBFqb1+ThduOKK3dEaQxHEcBHodBeYEbl0PfL1WRhRBCDE0S7Axjpmmya9cu/v73v2etmXPZZZdxxBFHUFxc3G0AkjIYq5b6I1l45vgyjj24hAt+8x8SpkV5vhuvy0i3d39URRZCCDE07dcE5SVLlnDOOedQXV2NpmksWLAg/VwikeA73/kOU6dOxefzUV1dzeWXX86OHTsyrjF79mw0Tcv4uvjiiwe5J4MvEAjwySefsGvXLhYuXJgOdPLy8rjsssu46667+OxnP0tJSQmapmFZilXbAiz+eBertgWwLJX1uqlApG3isWX1LPG4N/ojWXhdXRP1wSgVhR7y3I6MwKy7ROee6Ok9E0IIMbTt15Gd5uZmpk2bxpVXXskXvvCFjOfC4TArVqzgjjvuYNq0aTQ0NHDTTTdx7rnnsnz58oxzr776au6+++70Y6/XOyjt3x+i0Sj19fVEo1EA3G43F198MQ8//DAzZszgwgsvZOzYsRQWto689LZ432BtY7CvycIDWRU5+z3z8ZVjKznRL/k+QggxnOzXYGfOnDnMmTMn63N+v59FixZlHHvggQc49thj2bJlC6NGjUofz8vLo7KyckDbur+Zpsnu3bt59dVXGTVqVEbdnGnTpvH973+fSZMmUV5ejmG05tr0tXjfYK1a2pdk4YGqitzZPfuwton/eS2Ew+Pj+Anl3V9ICCHEkDCscnYCgYA9PdGuNsxTTz3Fk08+SUVFBXPmzOHOO++koCD70mOAWCxGLBZLPw4G7WkOpRRK9f9UReq6fb12MBhk7dq1PPXUU6xZs4YpU6bwta99LT1t43K5OO6448jLy0u/H6Rq5mygOZagqk3xPq+u4yl0teS0bOC4MSVZgxhNgykHtY4QWZbig62BrMHPvvaxLyZVFTBuhI8Pa5uoLHR3yC8KhOMcVlXApKqCHrerq3vmLXQRiUd4ZMmGnM4D2h/fy8EmfcwN0sfcsC997Olrhk2wE41G+e53v8ull16aMUVz2WWXMWbMGCorK1m9ejW33XYb77//fodRobbuvfdefvCDH3Q4HggEBizYCYVCAD1OFAaIx+Ps2rWLF198kX/+85/pmjkffPABa9asYcKECfj9fgoLC0kkEgQCgYzXb9gZIhQMMqHIwOXoWIQvv0gnFAzy3oYdjK/oeuPP97c28pd3t7GtIUzSVDgMjZHFeVxw1Eim1RT1uY/76ivHVvI/r4UIxyPkexzpUZimWJIxhQZfObaSpqae5+x0dc80FF63Tl0P79lwtb++l4NJ+pgbpI+5YV/6mBqs6M6wCHYSiQQXX3wxlmXx0EMPZTx39dVXp/8+ZcoUJkyYwNFHH82KFSuYPn161uvddttt3HLLLenHwWCQmpqadODQ31IBVE+WZQNYlsXu3btZtmwZf/rTn6irqwNA13WKioq48MILOeaYY6ioqMDl6nyKJrQzzvZmjREFTnTV8X0tS6c+bBHChb+LPJSlG3Zz10ubWqd1nDrNpsVb2yKs3rWJH82dymfGlfaqj/3lRL8fh8fHI0s2sqk+RMJK4tQ1xo7wc+2Jvc8v6uqeaSiKHIrt4WS392w46+3P63AkfcwN0sfcsC997On5Qz7YSSQSXHjhhWzevJlXX32122Bk+vTpOJ1O1q9f32mw43a7Mza9TEmt5hoIbVeLdSUYDLJ582aeeeaZdM2c1OtOPvlk5s6dS01NTY+CslKfG4ehEzMVniyFBKOmhUPXKfW5O22XZSkeWbKJpphJZaE3fZ5b16kotFdsPbJkEzPGlva4j/3t+AnlfGZcWb/kF3V3z+Km6vae5YL99b0cTNLH3CB9zA197WNOBDupQGf9+vW89tprlJaWdvuaNWvWkEgkqKqqGoQW9p94PM7OnTt54403ePrppwmHw+nnxowZw6WXXsrkyZM7JCB3pT9q5rTdPgINInGTpGUHSR6XnrG8e1TnaVIDrr+qInd3z5piScaO8Mvu6EIIMYzs12AnFAqxYcOG9OPNmzezcuVKSkpKqK6u5oILLmDFihX8/e9/xzTN9HROSUkJLpeLjRs38tRTT/G5z32OsrIy1q5dy6233sqRRx7J8ccfv7+61SuWZbFnzx4aGxtRSuF2u9OBjtfr5bzzzuOUU06hoqIinYDcU/1RvC+1vDtuWtQGosSSJkrZyctuh0FpvotEy/LuUQW9W/U0FHV1zwLhOGMKDa49UXZHF0KI4WS/BjvLly/npJNOSj9O5dHMmzePu+66ixdeeAGAI444IuN1r732GrNnz8blcvGvf/2LX/3qV4RCIWpqajjrrLO48847ezz6sT81NTWxa9cukslk+tjUqVPTU3Ff/OIXGTVqFKWlpX0evtzXmjkleS4spdjeYBctNHQNTQelIJow2d4Qwe91ppd3W5Zibe3w3mSzs3t2WFUBXzm2UrafEEKIYWa/BjuzZ8/ucvVTdyujampqWLx4cX83a8Clpqz+/e9/88EHHzBv3ryMYOaqq64iPz+/2wTkntqXmjkTKwswlcK0FC6Hht7STk0DdEU8qTCVYmJlAUs/3Mqjb3/IxvrmHhUvHMqy3bNJVQW9WtklhBBiaBjSOTu5RinFnj17WL9+PX/6059YvXo1AJMnT+aYY44B7BVXFRUV/b7Sp685LevqmjA0cBgaSQscukIDFNiPDQ1Dg2eXb+WFdzawOWhRlOfucfHCoaz9PcvlOhdCCJHLJNgZJOFwmJ07d7Jw4UIWLlyYrpkDdlL1McccQ0FBAeXl5Tgc+/fbYlkqPaLxya5mdE2n2u9lT3OMWNJK5+x4nfaqpHDc5Nl3tqDi9ootNHsVk0c3ZENOIYQQ+50EOwMskUiwc+dOli1bxnPPPZdOsga7psBFF13EMcccQ2VlJT6fbz+21NZ+TyiFoimaxOXQObjURzRhta7GcupEkxbETeqCUSYUOQgpjbbjH+035OyPFVNCCCFEb0iwM8C2bNnCo48+ytKlS9H11qXMp5xyCueccw5VVVWUlpaiZ6npMtiy7QkVS5oEIglqAxGchkaBxwnYyd+p5esVhW7qAxGchg7Jjtfdlw05hRBCiH0lwc4Aamxs5Fvf+hYNDQ3pY6maOePHj6eysjJrccP9wd4TaiOhWJLKtntCuRwcVORly94w2xsjjC7VcBtGxvL1S44dxcOvrSdhdtySAvq+IacQQgjRHyTYGUBFRUVMnTqVJUuW4PV6ueCCC5g1axZlZWUUFxfv7+ZlaFs8sP0y9wKPk0q/h91NcQLhJGjJjOXrM8aW8vKaWhoaAmguZ8tSLVtPixcKIYQQA0WCnQH2pS99CcuyOP3009Mblu7vBORsUsUDXUb26bRir4t40uJrJ49nTJmvw/L1a08cx68Wvs/mYAx/nqvXxQuFEEKIgTL0fuvmmKKiIq644go8Hg/V1dVDdm+TkjwXTkMjblp49I4FGWOmhcvQOWpUcdYk45njy0ieNJ5H365jY31zr4sXCiGEEANFgp0Blp+fT0lJCU1NTfu7KV3qj320ptUU8fikUaytbRrWFZSFEELklv2/BCjHeb3eIbHSqjupPaHy3fZO5pGEiWUpIgmTumCsx1NRqUJ8sw4pZ+pIvwQ6Qggh9ruh/1tYDJrUnlATqwoIx5LUh2KEY0kmVhUMywrIQgghBMg0lmhnX/bREkIIIYYiCXZEB33dR0sIIYQYimQaSwghhBA5TYIdIYQQQuQ0CXaEEEIIkdMk2BFCCCFETpNgRwghhBA5TYIdIYQQQuQ0CXaEEEIIkdMk2BFCCCFETpNgRwghhBA5TYIdIYQQQuQ0CXaEEEIIkdMk2BFCCCFETpNgRwghhBA5TYIdIYQQQuQ0CXaEEEIIkdMk2BFCCCFETpNgRwghhBA5TYIdIYQQQuQ0CXaEEEIIkdP2a7CzZMkSzjnnHKqrq9E0jQULFmQ8r5Tirrvuorq6Gq/Xy+zZs1mzZk3GObFYjBtvvJGysjJ8Ph/nnnsu27ZtG8ReZGdZilXbAiz+eBertgWwLLW/mySEEEIckPZrsNPc3My0adN48MEHsz5/3333cf/99/Pggw/yzjvvUFlZyWmnnUZTU1P6nJtuuon58+fzzDPP8OabbxIKhTj77LMxTXOwutHB0g27mffY21zzx+V888/vc92Ty7nzhTUs3bB7v7VJCCGEOFA59uebz5kzhzlz5mR9TinFL3/5S773ve9x/vnnA/DEE09QUVHB008/zTXXXEMgEOB3v/sdf/zjHzn11FMBePLJJ6mpqeGVV17hjDPOGLS+pCzdsJvb568iFEtSnOfCZegkTJPNu5v5/oJV/Pi8w5k5vmzQ2yWEEEIcqPZrsNOVzZs3U1dXx+mnn54+5na7mTVrFkuXLuWaa67h3XffJZFIZJxTXV3NlClTWLp0aafBTiwWIxaLpR8Hg0HADrCU6vt0k2UpHl68geZYgqpCD5qmAeDQdYqdTup3J3l48QaOG1OCrmt9fp+hKHXv9uX+DXUHQh/hwOin9DE3SB9zw770saevGbLBTl1dHQAVFRUZxysqKvj000/T57hcLoqLizuck3p9Nvfeey8/+MEPOhwPBAL79AO1YWeIUDDIhCIDl8NKH9dQ+B2K8UUGoWCQ9zbsYHxFfp/fZyhSShEKhQDSQV6uORD6CAdGP6WPuUH6mBv2pY+pwYruDNlgJ6V9x5VS3d6M7s657bbbuOWWW9KPg8EgNTU1+P1+CgsL+9zW0M4425s1RhQ40VXr+2soFNBk6uwMW4Rw4ff7+/w+Q1EqSPT7/Tn9DxJyu49wYPRT+pgbpI+5YV/62NPzh2ywU1lZCdijN1VVVenj9fX16dGeyspK4vE4DQ0NGaM79fX1zJw5s9Nru91u3G53h+Oapu3TD1Opz43D0ImZCo/ePvdbI2oqHLpOqc+dkz+0qfuXi31LORD6CAdGP6WPuUH6mBv62seenj9k6+yMGTOGyspKFi1alD4Wj8dZvHhxOpA56qijcDqdGefU1tayevXqLoOdgTK5upBxI/JpCCeyTIcpGsMJxo3IZ3J130ePhBBCCNE7+3VkJxQKsWHDhvTjzZs3s3LlSkpKShg1ahQ33XQT99xzDxMmTGDChAncc8895OXlcemllwL2kNdVV13FrbfeSmlpKSUlJXzzm99k6tSp6dVZg0nXNa6bNY7b56+iLhijKM+J29CJmya7Y3Hy3U6umzUu55KThRBCiKFsvwY7y5cv56STTko/TuXRzJs3j8cff5xvf/vbRCIRrr/+ehoaGjjuuOP45z//SUFBQfo1v/jFL3A4HFx44YVEIhFOOeUUHn/8cQzDGPT+AMwcX8Y9503l4cUb2VgfImApXDocVuXj4uMPlWXnQgghxCDTVC6vZ+uhYDCI3+8nEAjsU4JyW5alWLMjyN5wnGKvk5E+RXFxUc7OuSqlCAQCOZ9El+t9hAOjn9LH3CB9zA370see/v4esgnKw52ua0wdaa+4Sn0jhRBCCDH4hmyCshBCCCFEf5BgRwghhBA5TYIdIYQQQuQ0CXaEEEIIkdMk2BFCCCFETpNgRwghhBA5TYIdIYQQQuQ0CXaEEEIIkdMk2BFCCCFETpMKypDeoTwYDA7Y9YPBYJ+2rx8upI+540Dop/QxN0gfc8O+9DH1e7u7na8k2AGampoAqKmp2c8tEUIIIURvNTU14ff7O31eNgIFLMtix44dFBQUDEjkHAwGqampYevWrf220ehQI33MHQdCP6WPuUH6mBv2pY9KKZqamqiurkbXO8/MkZEdQNd1Ro4cOeDvU1hYmLM/rCnSx9xxIPRT+pgbpI+5oa997GpEJ0USlIUQQgiR0yTYEUIIIUROk2BnELjdbu68807cbvf+bsqAkT7mjgOhn9LH3CB9zA2D0UdJUBZCCCFETpORHSGEEELkNAl2hBBCCJHTJNgRQgghRE6TYEcIIYQQOU2CnT5asmQJ55xzDtXV1WiaxoIFC9LPJRIJvvOd7zB16lR8Ph/V1dVcfvnl7NixI+MasViMG2+8kbKyMnw+H+eeey7btm0b5J50rqs+tnfNNdegaRq//OUvM44P9T5Cz/q5bt06zj33XPx+PwUFBcyYMYMtW7aknx/q/eyuj6FQiBtuuIGRI0fi9XqZOHEiDz/8cMY5Q7mP9957L8cccwwFBQWMGDGCuXPn8tFHH2Wco5Tirrvuorq6Gq/Xy+zZs1mzZk3GOUO5j9B9P3Phs6cn38u2huNnT0/7OJw/d3rSx8H83JFgp4+am5uZNm0aDz74YIfnwuEwK1as4I477mDFihU8//zzfPzxx5x77rkZ5910003Mnz+fZ555hjfffJNQKMTZZ5+NaZqD1Y0uddXHthYsWMCyZcuorq7u8NxQ7yN038+NGzdywgkncNhhh/H666/z/vvvc8cdd+DxeNLnDPV+dtfHm2++mZdeeoknn3ySdevWcfPNN3PjjTfy17/+NX3OUO7j4sWL+drXvsZbb73FokWLSCaTnH766TQ3N6fPue+++7j//vt58MEHeeedd6isrOS0005L740HQ7uP0H0/c+Gzpyffy5Th+tnTkz4O98+dnvRxUD93lNhngJo/f36X57z99tsKUJ9++qlSSqnGxkbldDrVM888kz5n+/btStd19dJLLw1kc/uksz5u27ZNHXTQQWr16tVq9OjR6he/+EX6ueHWR6Wy9/Oiiy5SX/rSlzp9zXDrZ7Y+Tp48Wd19990Zx6ZPn66+//3vK6WGXx/r6+sVoBYvXqyUUsqyLFVZWal+8pOfpM+JRqPK7/erRx55RCk1/PqoVMd+ZjPcP3s662MuffZk62Oufe5k6+Ngfu7IyM4gCQQCaJpGUVERAO+++y6JRILTTz89fU51dTVTpkxh6dKl+6mVvWNZFl/+8pf51re+xeTJkzs8nyt9fPHFFznkkEM444wzGDFiBMcdd1zGNFAu9POEE07ghRdeYPv27SileO211/j4448544wzgOHXx0AgAEBJSQkAmzdvpq6uLqP9brebWbNmpds/3PoIHfvZ2TnD+bMnWx9z7bOnfR9z8XMn2/dxMD93JNgZBNFolO9+97tceuml6U3O6urqcLlcFBcXZ5xbUVFBXV3d/mhmr/30pz/F4XDw9a9/PevzudDH+vp6QqEQP/nJTzjzzDP55z//yXnnncf555/P4sWLgdzo569//WsmTZrEyJEjcblcnHnmmTz00EOccMIJwPDqo1KKW265hRNOOIEpU6YApNtYUVGRcW7b9g+nPkL2frY33D97OutjLn32ZOtjrn3udPZ9HMzPHdn1fIAlEgkuvvhiLMvioYce6vZ8pRSapg1Cy/bNu+++y69+9StWrFjR6/YOlz6C/X9YAJ///Oe5+eabATjiiCNYunQpjzzyCLNmzer0tcOpn7/+9a956623eOGFFxg9ejRLlizh+uuvp6qqilNPPbXT1w3FPt5www188MEHvPnmmx2ea9/WnrR/KPYRuu4n5MZnT7Y+5tpnT7Y+5trnTmc/q4P5uSMjOwMokUhw4YUXsnnzZhYtWpSxdX1lZSXxeJyGhoaM19TX13f4v8+h6I033qC+vp5Ro0bhcDhwOBx8+umn3HrrrRx88MHA8O8jQFlZGQ6Hg0mTJmUcnzhxYnpVxHDvZyQS4fbbb+f+++/nnHPO4fDDD+eGG27goosu4mc/+xkwfPp444038sILL/Daa68xcuTI9PHKykqADv832Lb9w6WP0Hk/U3Lhs6ezPubSZ09nfcylz53O+jjYnzsS7AyQ1IfN+vXreeWVVygtLc14/qijjsLpdLJo0aL0sdraWlavXs3MmTMHu7m99uUvf5kPPviAlStXpr+qq6v51re+xcsvvwwM/z4CuFwujjnmmA5LJj/++GNGjx4NDP9+JhIJEokEup75cWAYRvr/MId6H5VS3HDDDTz//PO8+uqrjBkzJuP5MWPGUFlZmdH+eDzO4sWL0+0f6n2E7vsJw/+zp7s+5sJnT3d9zIXPne76OOifO71KZxZpTU1N6r333lPvvfeeAtT999+v3nvvPfXpp5+qRCKhzj33XDVy5Ei1cuVKVVtbm/6KxWLpa1x77bVq5MiR6pVXXlErVqxQJ598spo2bZpKJpP7sWetuupjNu1XRCg19PuoVPf9fP7555XT6VT/+7//q9avX68eeOABZRiGeuONN9LXGOr97K6Ps2bNUpMnT1avvfaa2rRpk3rssceUx+NRDz30UPoaQ7mP1113nfL7/er111/P+PcWDofT5/zkJz9Rfr9fPf/882rVqlXqkksuUVVVVSoYDKbPGcp9VKr7fubCZ09PvpftDbfPnp70cbh/7vSkj4P5uSPBTh+99tprCujwNW/ePLV58+aszwHqtddeS18jEomoG264QZWUlCiv16vOPvtstWXLlv3XqXa66mM22T5whnoflepZP3/3u9+p8ePHK4/Ho6ZNm6YWLFiQcY2h3s/u+lhbW6uuuOIKVV1drTwejzr00EPVz3/+c2VZVvoaQ7mPnf17e+yxx9LnWJal7rzzTlVZWancbrc68cQT1apVqzKuM5T7qFT3/cyFz56efC/bG26fPT3t43D+3OlJHwfzc0draZQQQgghRE6SnB0hhBBC5DQJdoQQQgiR0yTYEUIIIUROk2BHCCGEEDlNgh0hhBBC5DQJdoQQQgiR0yTYEUIIIUROk2BHCNEjd90FRxzR+viKK2Du3MFvxyefgKbBypWD/95CiOFJgh0hhrErrrB/8WsaOJ0wdix885vQ3Dzw7/2rX8Hjj/fs3MEOUGbPbr0vbb+uvXZw3l8IMbQ49ncDhBD75swz4bHHIJGAN96Ar3zFDnYefrjjuYmEHRT1B7+/f64zUK6+Gu6+O/NYXt7Avmc8Di7XwL6HEKL3ZGRHiGHO7YbKSqipgUsvhcsugwUL7OdSU0+//7096uN2g1IQCMBXvwojRkBhIZx8Mrz/fuZ1f/ITqKiAggK46iqIRjOfbz+NZVnw05/C+PH2+4waBT/+sf1casPjI4+0R1hmz2593WOPwcSJ4PHAYYfBQw9lvs/bb9uv83jg6KPhvfd6dl/y8uz70varsNB+LjXS9PzzcNJJ9rnTpsF//pN5jaVL4cQTweu17+/Xv545anbwwfCjH9n3wu+3AyyA3/7WPj8vD847D+6/H4qKWt9b12H58sz3euABGD3a/v5kU1sLZ51lt2XMGHj6afv9f/nL1nPuvx+mTgWfz37/66+HUKj1+ccft9uxYAEccoh9T087DbZu7dk9FWK4kmBHiBzj9dojOCkbNsCf/wzPPdc6jXTWWVBXB//4B7z7LkyfDqecAnv32s//+c9w5512sLJ8OVRVdQxC2rvtNjvYueMOWLvW/mVcUWE/9/bb9p+vvGL/0n7+efvxb38L3/ue/T7r1sE999ivf+IJ+/nmZjj7bDj0ULudd91lT9P1l+99z77eypX2L/9LLoFk0n5u1So44ww4/3z44AN49ll480244YbMa/z3f8OUKXb77rgD/v1ve7rsG9+wr3vaaa1BH9gByqmn2kFeW4891jotmc3ll8OOHfD66/b38n//F+rrM8/Rdfj1r2H1avsevvoqfPvbmeeEw3Z7nnjCbmswCBdf3KvbJsTws+97mwoh9pd585T6/OdbHy9bplRpqVIXXmg/vvNOpZxOperrW8/517+UKixUKhrNvNa4cUr95jf23z/zGaWuvTbz+eOOU2ratOzvHQwq5XYr9dvfZm/n5s1KgVLvvZd5vKZGqaefzjz2wx/a76+U3Z6SEqWam1uff/jh7Ndqa9Ysu98+X+bX449ntufRR1tfs2aNfWzdOvvxl7+s1Fe/mnndN95QSteVikTsx6NHKzV3buY5F12k1FlnZR677DKl/P7Wx88+q1Rxcev3YOVKpTTNblc269bZbXvnndZj69fbx9pt9p3hz3+2fx5SHnvMfs1bb3W89rJlnV9HiOFORnaEGOb+/nfIz7enJD7zGXva5YEHWp8fPRrKy1sfv/uuPbVRWmq/LvW1eTNs3Gifs26dfa222j9ua906iMXs0aGe2rXLnj656qrMdvzoR5ntmDYtM9emq3a0ddll9shK26/zzss85/DDW/9eVWX/mRotefdde9qnbdvOOMOertu8ufV1Rx+dec2PPoJjj8081v7x3LngcMD8+fbj3//enk47+ODsffnoI/v86dNbj40fD8XFmee99po9knTQQfb04+WXw549mVNvDkdmmw87zJ7aWrcu+3sLkQskQVmIYe6kk+xkZKcTqqs7JiD7fJmPLcv+xf766x2vlcor6S2vt/evsSz7z9/+Fo47LvM5w7D/7Cx/pSf8fjsg6Erbe5WaPkq1y7LgmmvsPJ32Ro1q/Xv7+6tUx6mo9v1wueDLX7anrs4/357ya5t7015n96Ht8U8/hc99zp5C++EPoaTEnna76qrMaU3IPlXW2fSZELlAgh0hhjmfr/tf6m1Nn27n6zgcnY8kTJwIb71ljwykvPVW59ecMMEOeP71L3s1WHupFUqm2XqsosIegdi0yR6FyWbSJPjjHyESaQ2oumpHf5o+Hdas6d29BXukJJWjlNI+GRns+zRlip0LlUjYQU9X10wm7eTso46yj23YAI2Nme+RTMLPf27n7oCde9VeMmmfmxpt+ugj+zqHHdbTHgox/Mg0lhAHmFNPtaeC5s6Fl1+2VwctXQrf/37rL+VvfMOeWvn97+Hjj+1k5TVrOr+mxwPf+Y6dDPuHP9jTUG+9Bb/7nf38iBF2sPLSS7Bzp70aDOyE43vvtWv2fPyxnRT82GP2qiKwV5fpuj06sXatnVD9s5/1rJ/hsB3Utf1qaOj5ffrOd+zVWV/7mj0Ftn49vPAC3Hhj16+78Ua7nfffb7/mN7+BhQs7jpxMnAgzZtjvc8klXY+OHXaY/X376lftQOq99+y/e72t1x03zg5kHnjADiD/+Ed45JGO13I67TYuWwYrVsCVV9rtaD/VJkQukWBHiAOMptm/jE88Ef7rv+xVSBdfbAc9qdVTF10E/+//2b+IjzrKniK57rqur3vHHXDrrfbrJk60r5HKf3E47FVCv/mNPdX2+c/bx7/yFXj0UTs3ZupUmDXL/ntqqXp+Pvztb3agc+SR9uqpn/60Z/387W/t6bq2X5dc0vP7dPjhsHixHbB89rP2+99xR2tuT2eOP94OMu6/3843eukluPlmOyBs76qr7No8//Vf3bfnD3+wvz8nnmjnHl19tZ2Xk7ruEUfY7/nTn9ojRk89ZQeS7eXl2d/XSy+1g16vF555pvv3F2I405Tal1lxIYQQ3bn6avjwQ7voY1s//rEdaKxa1ftrbttm19J55ZWeJ4Y//jjcdFPm9JcQBwLJ2RFCiH72s5/Zq6J8PnsK64knMusUhUL26qcHHrCTiXvi1Vft102datcq+va37ZyrE08ckC4IkVMk2BFCiH729ttw333Q1GRXrv71rzMTt2+4Af70JztvqidTWGAnMd9+u52PU1AAM2faU1X9tf2HELlMprGEEEIIkdMkQVkIIYQQOU2CHSGEEELkNAl2hBBCCJHTJNgRQgghRE6TYEcIIYQQOU2CHSGEEELkNAl2hBBCCJHTJNgRQgghRE6TYEcIIYQQOe3/A/YUWLEX8U17AAAAAElFTkSuQmCC",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "sn.regplot(x=ave_predict , y=y_test,line_kws={\"lw\":2,'ls':'--','color':'black',\"alpha\":0.7})\n",
        "plt.xlabel('Predicted Energy gap', color='blue')\n",
        "plt.ylabel('Observed Energy gap', color ='blue')\n",
        "plt.title(\"Test set\", color='red')\n",
        "plt.grid(alpha=0.2)\n",
        "R2 = mpatches.Patch(label=\"R2={:04.2f}\".format(R2))\n",
        "MAE = mpatches.Patch(label=\"MAE={:04.2f}\".format(MAE))\n",
        "plt.legend(handles=[R2, MAE])\n",
        "\n",
        "# Save the plot as an SVG image\n",
        "plt.savefig('plot.svg', format='svg')\n",
        "\n",
        "# Show the plot (optional)\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "gtASLn2-dlvF",
      "metadata": {
        "id": "gtASLn2-dlvF"
      },
      "source": [
        "## Saving model for future use"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 55,
      "id": "vTX4x89xdlvF",
      "metadata": {
        "id": "vTX4x89xdlvF"
      },
      "outputs": [],
      "source": [
        "from sklearn.preprocessing import StandardScaler\n",
        "\n",
        "scaler = StandardScaler()\n",
        "scaler.fit(X_train)\n",
        "\n",
        "import pickle\n",
        "with open('lgbm_model.pkl','wb') as f:\n",
        "          pickle.dump(lgbm_opt,f)\n",
        "        \n",
        "with open('hgbr_model.pkl','wb') as f:\n",
        "          pickle.dump(gbr_opt,f)\n",
        "with open('scaler.pkl','wb') as f:\n",
        "          pickle.dump(scaler,f)"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "PEzCtxUQdlvF",
      "metadata": {
        "id": "PEzCtxUQdlvF"
      },
      "source": [
        "##  Loading the model"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 56,
      "id": "mCmtXcSmdlvF",
      "metadata": {
        "id": "mCmtXcSmdlvF"
      },
      "outputs": [],
      "source": [
        "# with open('lgbm_model.pkl','rb') as f:\n",
        "#     model_lgbm = pickle.load(f)\n",
        "        \n",
        "# with open('hgbr_model.pkl','rb') as f:\n",
        "#     model_hgbr = pickle.load(f)"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "include_colab_link": true,
      "name": "HOMO-LUMO Energy Gap Prediction using Molecular descriptors.ipynb",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.11.3"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 5
}