HugoHakem_FinalProject_Part_2_Classification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h1><center> BioE245 Final project Part2: Classification </h1></center>\n",
    "<h2><center> Efficacy Comparison of various model for Peripheral blood mononuclear cells classification using transcriptomics data </h2></center>\n",
    "    \n",
    "## Hugo Hakem Meng BioE 23-24'\n",
    "\n",
    "### Acknowledgement\n",
    "This work stands as the final project for the BioE245 class taught in Spring2024 by Professor. Liana Lareau, and GSIs. Michael Herschl, Amy Lu.\n",
    "\n",
    "# 0) Some background and motivation\n",
    "Peripheral blood mononuclear cells (PBMCs) are a group of different cell types in your blood, a subset of what we call white blood cells. They include many important parts of your immune system: T cells, B cells, natural killer cells, and so on. These cells are responsible for your innate and adaptive immune responses. When you get infected or vaccinated, they kick into high gear. In different situations, or with different diseases, the mix of different cell types shifts and so does the gene expression within one type of cell. PBMCs can be isolated very easily from patient blood samples, just by spinning the blood in a centrifuge that separates different cell types by weight. So, looking at these cells can be very informative.\n",
    "\n",
    "For this present work, two dataset has been furnished: \n",
    "+ labels.csv stores the association between a cell barcode and a cell type label.\n",
    "+ processed_counts.csv stores the normalized log read counts for each cell, where each row represents a single cell, and each column represents a gene.\n",
    "    \n",
    "The goal of this part of the project is to classify the cell using supervised machine larning method. \n",
    "\n",
    "This Notebook is partitionned as follow.\n",
    "1. [Data Preprocessing](#PrePro)\n",
    "2. [Model Implementation](#Model)\n",
    "   1. [Decision Trees](#DecTrees)\n",
    "        + [XGBoost](#XGBoost)\n",
    "        + [RandomForest](#RandomForest)\n",
    "   2. [Bayes Model](#Bayes)\n",
    "        + [Gaussian Naive Bayes](#Gaussian)\n",
    "        + [Bernouilli Naive Bayes](#Bernouilli)\n",
    "        + [Multivariate Gaussian Bayes](#MultiGaussian)\n",
    "   3. [Feed Forward Neural Network](#FFNN)\n",
    "3. [Conclusion](#Conclusion)\n",
    "\n",
    "\n",
    "To run the present Notebook, here is a list of command to run sequentially in your environment and which will install the package required. \n",
    "\n",
    "```conda install ipykernel conda-forge -y```\n",
    "```conda install pytorch torchvision torchaudio pytorch-cuda=11.8 -c pytorch -c nvidia```\n",
    "```conda install pandas matplotlib seaborn tqdm -c conda-forge -y```\n",
    "```conda install -c conda-forge py-xgboost```\n",
    "    \n",
    "**Please Note: for each model recquiring some fine-tuning, several parameters has been tested and only the best architecture are displayed** "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1) Data Preprocessing <a id=\"PrePro\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 371,
   "metadata": {
    "id": "y_Dn9rzyiqxi"
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from xgboost import XGBClassifier\n",
    "\n",
    "\n",
    "from sklearn.metrics import roc_auc_score, accuracy_score\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.naive_bayes import GaussianNB, BernoulliNB, MultinomialNB, ComplementNB\n",
    "from sklearn.mixture import GaussianMixture\n",
    "from scipy.stats import multivariate_normal\n",
    "\n",
    "from tqdm import tqdm\n",
    "\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "from torch.utils.data import Dataset, DataLoader"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 289,
   "metadata": {
    "id": "WelsjSzviy4m"
   },
   "outputs": [],
   "source": [
    "labels_pd = pd.read_csv(\"labels.csv\")\n",
    "counts_pd = pd.read_csv(\"processed_counts.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "GGLKx-6qiz-_"
   },
   "source": [
    "This data associates a cell barcode, such as \"AAAGCCTGGCTAAC-1\", to a certain cell type label, such as \"CD14+ Monocyte\". For each cell barcode, there are also log RNA seq counts of 765 different genes, such as HES4.\n",
    "\n",
    "label.csv stores the association between a cell barcode and a cell type label.\n",
    "\n",
    "processed_counts.csv stores the normalized log read counts for each cell, where each row represents a single cell, and each column represents a gene."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 290,
   "metadata": {},
   "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>index</th>\n",
       "      <th>bulk_labels</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>AAAGCCTGGCTAAC-1</td>\n",
       "      <td>CD14+ Monocyte</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>AAATTCGATGCACA-1</td>\n",
       "      <td>Dendritic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>AACACGTGGTCTTT-1</td>\n",
       "      <td>CD56+ NK</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>AAGTGCACGTGCTA-1</td>\n",
       "      <td>CD4+/CD25 T Reg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ACACGAACGGAGTG-1</td>\n",
       "      <td>Dendritic</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              index      bulk_labels\n",
       "0  AAAGCCTGGCTAAC-1   CD14+ Monocyte\n",
       "1  AAATTCGATGCACA-1        Dendritic\n",
       "2  AACACGTGGTCTTT-1         CD56+ NK\n",
       "3  AAGTGCACGTGCTA-1  CD4+/CD25 T Reg\n",
       "4  ACACGAACGGAGTG-1        Dendritic"
      ]
     },
     "execution_count": 290,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labels_pd.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 291,
   "metadata": {},
   "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>Unnamed: 0</th>\n",
       "      <th>HES4</th>\n",
       "      <th>TNFRSF4</th>\n",
       "      <th>SSU72</th>\n",
       "      <th>PARK7</th>\n",
       "      <th>RBP7</th>\n",
       "      <th>SRM</th>\n",
       "      <th>MAD2L2</th>\n",
       "      <th>AGTRAP</th>\n",
       "      <th>TNFRSF1B</th>\n",
       "      <th>...</th>\n",
       "      <th>ATP5O</th>\n",
       "      <th>MRPS6</th>\n",
       "      <th>TTC3</th>\n",
       "      <th>U2AF1</th>\n",
       "      <th>CSTB</th>\n",
       "      <th>SUMO3</th>\n",
       "      <th>ITGB2</th>\n",
       "      <th>S100B</th>\n",
       "      <th>PRMT2</th>\n",
       "      <th>MT-ND3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>AAAGCCTGGCTAAC-1</td>\n",
       "      <td>-0.326</td>\n",
       "      <td>-0.191</td>\n",
       "      <td>-0.728</td>\n",
       "      <td>-0.301</td>\n",
       "      <td>3.386</td>\n",
       "      <td>-0.531</td>\n",
       "      <td>2.016</td>\n",
       "      <td>3.377</td>\n",
       "      <td>4.841</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.146</td>\n",
       "      <td>-0.532</td>\n",
       "      <td>-0.341</td>\n",
       "      <td>0.303</td>\n",
       "      <td>1.404</td>\n",
       "      <td>4.294</td>\n",
       "      <td>0.519</td>\n",
       "      <td>-0.21</td>\n",
       "      <td>-0.636</td>\n",
       "      <td>4.011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>AAATTCGATGCACA-1</td>\n",
       "      <td>1.171</td>\n",
       "      <td>-0.191</td>\n",
       "      <td>0.795</td>\n",
       "      <td>-1.200</td>\n",
       "      <td>-0.174</td>\n",
       "      <td>-0.531</td>\n",
       "      <td>1.889</td>\n",
       "      <td>-0.486</td>\n",
       "      <td>-0.459</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.136</td>\n",
       "      <td>-0.532</td>\n",
       "      <td>-0.341</td>\n",
       "      <td>-0.905</td>\n",
       "      <td>2.849</td>\n",
       "      <td>-0.585</td>\n",
       "      <td>1.172</td>\n",
       "      <td>-0.21</td>\n",
       "      <td>2.630</td>\n",
       "      <td>-0.490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>AACACGTGGTCTTT-1</td>\n",
       "      <td>-0.326</td>\n",
       "      <td>-0.191</td>\n",
       "      <td>0.483</td>\n",
       "      <td>-1.200</td>\n",
       "      <td>-0.174</td>\n",
       "      <td>-0.531</td>\n",
       "      <td>-0.451</td>\n",
       "      <td>0.971</td>\n",
       "      <td>-0.459</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.136</td>\n",
       "      <td>2.606</td>\n",
       "      <td>-0.341</td>\n",
       "      <td>-0.905</td>\n",
       "      <td>-0.455</td>\n",
       "      <td>-0.585</td>\n",
       "      <td>0.722</td>\n",
       "      <td>-0.21</td>\n",
       "      <td>0.663</td>\n",
       "      <td>-0.490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>AAGTGCACGTGCTA-1</td>\n",
       "      <td>-0.326</td>\n",
       "      <td>-0.191</td>\n",
       "      <td>1.134</td>\n",
       "      <td>-0.157</td>\n",
       "      <td>-0.174</td>\n",
       "      <td>-0.531</td>\n",
       "      <td>-0.451</td>\n",
       "      <td>-0.486</td>\n",
       "      <td>-0.459</td>\n",
       "      <td>...</td>\n",
       "      <td>1.161</td>\n",
       "      <td>-0.532</td>\n",
       "      <td>-0.341</td>\n",
       "      <td>-0.905</td>\n",
       "      <td>-0.119</td>\n",
       "      <td>-0.585</td>\n",
       "      <td>0.766</td>\n",
       "      <td>-0.21</td>\n",
       "      <td>-0.636</td>\n",
       "      <td>-0.490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ACACGAACGGAGTG-1</td>\n",
       "      <td>-0.326</td>\n",
       "      <td>-0.191</td>\n",
       "      <td>-0.728</td>\n",
       "      <td>-0.607</td>\n",
       "      <td>-0.174</td>\n",
       "      <td>-0.531</td>\n",
       "      <td>-0.451</td>\n",
       "      <td>0.787</td>\n",
       "      <td>-0.459</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.136</td>\n",
       "      <td>0.839</td>\n",
       "      <td>1.679</td>\n",
       "      <td>-0.108</td>\n",
       "      <td>-0.534</td>\n",
       "      <td>-0.585</td>\n",
       "      <td>-0.007</td>\n",
       "      <td>-0.21</td>\n",
       "      <td>-0.636</td>\n",
       "      <td>-0.490</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 766 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         Unnamed: 0   HES4  TNFRSF4  SSU72  PARK7   RBP7    SRM  MAD2L2  \\\n",
       "0  AAAGCCTGGCTAAC-1 -0.326   -0.191 -0.728 -0.301  3.386 -0.531   2.016   \n",
       "1  AAATTCGATGCACA-1  1.171   -0.191  0.795 -1.200 -0.174 -0.531   1.889   \n",
       "2  AACACGTGGTCTTT-1 -0.326   -0.191  0.483 -1.200 -0.174 -0.531  -0.451   \n",
       "3  AAGTGCACGTGCTA-1 -0.326   -0.191  1.134 -0.157 -0.174 -0.531  -0.451   \n",
       "4  ACACGAACGGAGTG-1 -0.326   -0.191 -0.728 -0.607 -0.174 -0.531  -0.451   \n",
       "\n",
       "   AGTRAP  TNFRSF1B  ...  ATP5O  MRPS6   TTC3  U2AF1   CSTB  SUMO3  ITGB2  \\\n",
       "0   3.377     4.841  ... -0.146 -0.532 -0.341  0.303  1.404  4.294  0.519   \n",
       "1  -0.486    -0.459  ... -1.136 -0.532 -0.341 -0.905  2.849 -0.585  1.172   \n",
       "2   0.971    -0.459  ... -1.136  2.606 -0.341 -0.905 -0.455 -0.585  0.722   \n",
       "3  -0.486    -0.459  ...  1.161 -0.532 -0.341 -0.905 -0.119 -0.585  0.766   \n",
       "4   0.787    -0.459  ... -1.136  0.839  1.679 -0.108 -0.534 -0.585 -0.007   \n",
       "\n",
       "   S100B  PRMT2  MT-ND3  \n",
       "0  -0.21 -0.636   4.011  \n",
       "1  -0.21  2.630  -0.490  \n",
       "2  -0.21  0.663  -0.490  \n",
       "3  -0.21 -0.636  -0.490  \n",
       "4  -0.21 -0.636  -0.490  \n",
       "\n",
       "[5 rows x 766 columns]"
      ]
     },
     "execution_count": 291,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counts_pd.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Let's check the class imbalance of the data set**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 292,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "bulk_labels\n",
       "CD14+ Monocyte                  18.428571\n",
       "CD19+ B                         13.571429\n",
       "CD34+                            1.857143\n",
       "CD4+/CD25 T Reg                  9.714286\n",
       "CD4+/CD45RA+/CD25- Naive T       1.142857\n",
       "CD4+/CD45RO+ Memory              2.714286\n",
       "CD56+ NK                         4.428571\n",
       "CD8+ Cytotoxic T                 7.714286\n",
       "CD8+/CD45RA+ Naive Cytotoxic     6.142857\n",
       "Dendritic                       34.285714\n",
       "dtype: float64"
      ]
     },
     "execution_count": 292,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labels_pd.groupby(\"bulk_labels\").size() * 100 / labels_pd.shape[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this data set, **not all class are equivalent in term of proportion** which means that specific attention must be given to the classifier architecture to deal with this imbalance. Otherwise the classifier will be biased to predict the class with larger population proportion. **For this reason weight will be associated to each class.** Note however that using class weight will not necessarly lead to classification improvement as the **imbalance remain moderate.** Indeed, a data set is commonly said to be severely imbalanced for class proportion less than 1%. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 293,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.0620155 , 0.03333333, 0.25806452, 0.11764706, 0.08421053,\n",
       "       0.14814815, 0.42105263, 0.18604651, 1.        , 0.61538462])"
      ]
     },
     "execution_count": 293,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "class_perc = labels_pd.groupby(\"bulk_labels\").size() / labels_pd.shape[0]\n",
    "class_weight = class_perc.min() / class_perc\n",
    "class_weight = class_weight[categories].values\n",
    "class_weight"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 294,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 592
    },
    "id": "aIX8kcTXi7EV",
    "outputId": "aacbc984-4c80-42bd-d4d5-01bd54f12ca8"
   },
   "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>HES4</th>\n",
       "      <th>TNFRSF4</th>\n",
       "      <th>SSU72</th>\n",
       "      <th>PARK7</th>\n",
       "      <th>RBP7</th>\n",
       "      <th>SRM</th>\n",
       "      <th>MAD2L2</th>\n",
       "      <th>AGTRAP</th>\n",
       "      <th>TNFRSF1B</th>\n",
       "      <th>EFHD2</th>\n",
       "      <th>...</th>\n",
       "      <th>MRPS6</th>\n",
       "      <th>TTC3</th>\n",
       "      <th>U2AF1</th>\n",
       "      <th>CSTB</th>\n",
       "      <th>SUMO3</th>\n",
       "      <th>ITGB2</th>\n",
       "      <th>S100B</th>\n",
       "      <th>PRMT2</th>\n",
       "      <th>MT-ND3</th>\n",
       "      <th>bulk_labels</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>AAAGCCTGGCTAAC-1</th>\n",
       "      <td>-0.326</td>\n",
       "      <td>-0.191</td>\n",
       "      <td>-0.728</td>\n",
       "      <td>-0.301</td>\n",
       "      <td>3.386</td>\n",
       "      <td>-0.531</td>\n",
       "      <td>2.016</td>\n",
       "      <td>3.377</td>\n",
       "      <td>4.841</td>\n",
       "      <td>-0.525</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.532</td>\n",
       "      <td>-0.341</td>\n",
       "      <td>0.303</td>\n",
       "      <td>1.404</td>\n",
       "      <td>4.294</td>\n",
       "      <td>0.519</td>\n",
       "      <td>-0.21</td>\n",
       "      <td>-0.636</td>\n",
       "      <td>4.011</td>\n",
       "      <td>CD14+ Monocyte</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AAATTCGATGCACA-1</th>\n",
       "      <td>1.171</td>\n",
       "      <td>-0.191</td>\n",
       "      <td>0.795</td>\n",
       "      <td>-1.200</td>\n",
       "      <td>-0.174</td>\n",
       "      <td>-0.531</td>\n",
       "      <td>1.889</td>\n",
       "      <td>-0.486</td>\n",
       "      <td>-0.459</td>\n",
       "      <td>-0.525</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.532</td>\n",
       "      <td>-0.341</td>\n",
       "      <td>-0.905</td>\n",
       "      <td>2.849</td>\n",
       "      <td>-0.585</td>\n",
       "      <td>1.172</td>\n",
       "      <td>-0.21</td>\n",
       "      <td>2.630</td>\n",
       "      <td>-0.490</td>\n",
       "      <td>Dendritic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AACACGTGGTCTTT-1</th>\n",
       "      <td>-0.326</td>\n",
       "      <td>-0.191</td>\n",
       "      <td>0.483</td>\n",
       "      <td>-1.200</td>\n",
       "      <td>-0.174</td>\n",
       "      <td>-0.531</td>\n",
       "      <td>-0.451</td>\n",
       "      <td>0.971</td>\n",
       "      <td>-0.459</td>\n",
       "      <td>2.286</td>\n",
       "      <td>...</td>\n",
       "      <td>2.606</td>\n",
       "      <td>-0.341</td>\n",
       "      <td>-0.905</td>\n",
       "      <td>-0.455</td>\n",
       "      <td>-0.585</td>\n",
       "      <td>0.722</td>\n",
       "      <td>-0.21</td>\n",
       "      <td>0.663</td>\n",
       "      <td>-0.490</td>\n",
       "      <td>CD56+ NK</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AAGTGCACGTGCTA-1</th>\n",
       "      <td>-0.326</td>\n",
       "      <td>-0.191</td>\n",
       "      <td>1.134</td>\n",
       "      <td>-0.157</td>\n",
       "      <td>-0.174</td>\n",
       "      <td>-0.531</td>\n",
       "      <td>-0.451</td>\n",
       "      <td>-0.486</td>\n",
       "      <td>-0.459</td>\n",
       "      <td>-0.525</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.532</td>\n",
       "      <td>-0.341</td>\n",
       "      <td>-0.905</td>\n",
       "      <td>-0.119</td>\n",
       "      <td>-0.585</td>\n",
       "      <td>0.766</td>\n",
       "      <td>-0.21</td>\n",
       "      <td>-0.636</td>\n",
       "      <td>-0.490</td>\n",
       "      <td>CD4+/CD25 T Reg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ACACGAACGGAGTG-1</th>\n",
       "      <td>-0.326</td>\n",
       "      <td>-0.191</td>\n",
       "      <td>-0.728</td>\n",
       "      <td>-0.607</td>\n",
       "      <td>-0.174</td>\n",
       "      <td>-0.531</td>\n",
       "      <td>-0.451</td>\n",
       "      <td>0.787</td>\n",
       "      <td>-0.459</td>\n",
       "      <td>1.932</td>\n",
       "      <td>...</td>\n",
       "      <td>0.839</td>\n",
       "      <td>1.679</td>\n",
       "      <td>-0.108</td>\n",
       "      <td>-0.534</td>\n",
       "      <td>-0.585</td>\n",
       "      <td>-0.007</td>\n",
       "      <td>-0.21</td>\n",
       "      <td>-0.636</td>\n",
       "      <td>-0.490</td>\n",
       "      <td>Dendritic</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>TGGCACCTCCAACA-8</th>\n",
       "      <td>-0.326</td>\n",
       "      <td>-0.191</td>\n",
       "      <td>0.372</td>\n",
       "      <td>-0.584</td>\n",
       "      <td>-0.174</td>\n",
       "      <td>0.509</td>\n",
       "      <td>-0.451</td>\n",
       "      <td>2.160</td>\n",
       "      <td>-0.459</td>\n",
       "      <td>0.751</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.532</td>\n",
       "      <td>1.758</td>\n",
       "      <td>1.578</td>\n",
       "      <td>0.055</td>\n",
       "      <td>-0.585</td>\n",
       "      <td>0.561</td>\n",
       "      <td>-0.21</td>\n",
       "      <td>0.543</td>\n",
       "      <td>2.593</td>\n",
       "      <td>Dendritic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TGTGAGTGCTTTAC-8</th>\n",
       "      <td>3.166</td>\n",
       "      <td>-0.191</td>\n",
       "      <td>-0.728</td>\n",
       "      <td>-1.200</td>\n",
       "      <td>-0.174</td>\n",
       "      <td>-0.531</td>\n",
       "      <td>-0.451</td>\n",
       "      <td>-0.486</td>\n",
       "      <td>-0.459</td>\n",
       "      <td>1.536</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.532</td>\n",
       "      <td>-0.341</td>\n",
       "      <td>0.432</td>\n",
       "      <td>0.753</td>\n",
       "      <td>-0.585</td>\n",
       "      <td>-0.171</td>\n",
       "      <td>-0.21</td>\n",
       "      <td>1.268</td>\n",
       "      <td>-0.490</td>\n",
       "      <td>Dendritic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TGTTACTGGCGATT-8</th>\n",
       "      <td>-0.326</td>\n",
       "      <td>-0.191</td>\n",
       "      <td>-0.728</td>\n",
       "      <td>-1.200</td>\n",
       "      <td>-0.174</td>\n",
       "      <td>1.784</td>\n",
       "      <td>-0.451</td>\n",
       "      <td>-0.486</td>\n",
       "      <td>-0.459</td>\n",
       "      <td>-0.525</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.532</td>\n",
       "      <td>-0.341</td>\n",
       "      <td>1.858</td>\n",
       "      <td>0.814</td>\n",
       "      <td>-0.585</td>\n",
       "      <td>0.152</td>\n",
       "      <td>-0.21</td>\n",
       "      <td>-0.636</td>\n",
       "      <td>1.226</td>\n",
       "      <td>CD4+/CD25 T Reg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TTCAGTACCGGGAA-8</th>\n",
       "      <td>-0.326</td>\n",
       "      <td>-0.191</td>\n",
       "      <td>-0.728</td>\n",
       "      <td>-0.386</td>\n",
       "      <td>-0.174</td>\n",
       "      <td>-0.531</td>\n",
       "      <td>-0.451</td>\n",
       "      <td>-0.486</td>\n",
       "      <td>-0.459</td>\n",
       "      <td>-0.525</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.532</td>\n",
       "      <td>-0.341</td>\n",
       "      <td>2.378</td>\n",
       "      <td>0.420</td>\n",
       "      <td>0.888</td>\n",
       "      <td>-0.326</td>\n",
       "      <td>-0.21</td>\n",
       "      <td>-0.636</td>\n",
       "      <td>-0.490</td>\n",
       "      <td>CD19+ B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TTGAGGTGGAGAGC-8</th>\n",
       "      <td>-0.326</td>\n",
       "      <td>-0.191</td>\n",
       "      <td>0.148</td>\n",
       "      <td>0.762</td>\n",
       "      <td>-0.174</td>\n",
       "      <td>0.298</td>\n",
       "      <td>-0.451</td>\n",
       "      <td>-0.486</td>\n",
       "      <td>0.505</td>\n",
       "      <td>-0.525</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.532</td>\n",
       "      <td>-0.341</td>\n",
       "      <td>1.073</td>\n",
       "      <td>-0.628</td>\n",
       "      <td>-0.585</td>\n",
       "      <td>0.239</td>\n",
       "      <td>-0.21</td>\n",
       "      <td>-0.636</td>\n",
       "      <td>-0.490</td>\n",
       "      <td>Dendritic</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>700 rows × 766 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                   HES4  TNFRSF4  SSU72  PARK7   RBP7    SRM  MAD2L2  AGTRAP  \\\n",
       "AAAGCCTGGCTAAC-1 -0.326   -0.191 -0.728 -0.301  3.386 -0.531   2.016   3.377   \n",
       "AAATTCGATGCACA-1  1.171   -0.191  0.795 -1.200 -0.174 -0.531   1.889  -0.486   \n",
       "AACACGTGGTCTTT-1 -0.326   -0.191  0.483 -1.200 -0.174 -0.531  -0.451   0.971   \n",
       "AAGTGCACGTGCTA-1 -0.326   -0.191  1.134 -0.157 -0.174 -0.531  -0.451  -0.486   \n",
       "ACACGAACGGAGTG-1 -0.326   -0.191 -0.728 -0.607 -0.174 -0.531  -0.451   0.787   \n",
       "...                 ...      ...    ...    ...    ...    ...     ...     ...   \n",
       "TGGCACCTCCAACA-8 -0.326   -0.191  0.372 -0.584 -0.174  0.509  -0.451   2.160   \n",
       "TGTGAGTGCTTTAC-8  3.166   -0.191 -0.728 -1.200 -0.174 -0.531  -0.451  -0.486   \n",
       "TGTTACTGGCGATT-8 -0.326   -0.191 -0.728 -1.200 -0.174  1.784  -0.451  -0.486   \n",
       "TTCAGTACCGGGAA-8 -0.326   -0.191 -0.728 -0.386 -0.174 -0.531  -0.451  -0.486   \n",
       "TTGAGGTGGAGAGC-8 -0.326   -0.191  0.148  0.762 -0.174  0.298  -0.451  -0.486   \n",
       "\n",
       "                  TNFRSF1B  EFHD2  ...  MRPS6   TTC3  U2AF1   CSTB  SUMO3  \\\n",
       "AAAGCCTGGCTAAC-1     4.841 -0.525  ... -0.532 -0.341  0.303  1.404  4.294   \n",
       "AAATTCGATGCACA-1    -0.459 -0.525  ... -0.532 -0.341 -0.905  2.849 -0.585   \n",
       "AACACGTGGTCTTT-1    -0.459  2.286  ...  2.606 -0.341 -0.905 -0.455 -0.585   \n",
       "AAGTGCACGTGCTA-1    -0.459 -0.525  ... -0.532 -0.341 -0.905 -0.119 -0.585   \n",
       "ACACGAACGGAGTG-1    -0.459  1.932  ...  0.839  1.679 -0.108 -0.534 -0.585   \n",
       "...                    ...    ...  ...    ...    ...    ...    ...    ...   \n",
       "TGGCACCTCCAACA-8    -0.459  0.751  ... -0.532  1.758  1.578  0.055 -0.585   \n",
       "TGTGAGTGCTTTAC-8    -0.459  1.536  ... -0.532 -0.341  0.432  0.753 -0.585   \n",
       "TGTTACTGGCGATT-8    -0.459 -0.525  ... -0.532 -0.341  1.858  0.814 -0.585   \n",
       "TTCAGTACCGGGAA-8    -0.459 -0.525  ... -0.532 -0.341  2.378  0.420  0.888   \n",
       "TTGAGGTGGAGAGC-8     0.505 -0.525  ... -0.532 -0.341  1.073 -0.628 -0.585   \n",
       "\n",
       "                  ITGB2  S100B  PRMT2  MT-ND3      bulk_labels  \n",
       "AAAGCCTGGCTAAC-1  0.519  -0.21 -0.636   4.011   CD14+ Monocyte  \n",
       "AAATTCGATGCACA-1  1.172  -0.21  2.630  -0.490        Dendritic  \n",
       "AACACGTGGTCTTT-1  0.722  -0.21  0.663  -0.490         CD56+ NK  \n",
       "AAGTGCACGTGCTA-1  0.766  -0.21 -0.636  -0.490  CD4+/CD25 T Reg  \n",
       "ACACGAACGGAGTG-1 -0.007  -0.21 -0.636  -0.490        Dendritic  \n",
       "...                 ...    ...    ...     ...              ...  \n",
       "TGGCACCTCCAACA-8  0.561  -0.21  0.543   2.593        Dendritic  \n",
       "TGTGAGTGCTTTAC-8 -0.171  -0.21  1.268  -0.490        Dendritic  \n",
       "TGTTACTGGCGATT-8  0.152  -0.21 -0.636   1.226  CD4+/CD25 T Reg  \n",
       "TTCAGTACCGGGAA-8 -0.326  -0.21 -0.636  -0.490          CD19+ B  \n",
       "TTGAGGTGGAGAGC-8  0.239  -0.21 -0.636  -0.490        Dendritic  \n",
       "\n",
       "[700 rows x 766 columns]"
      ]
     },
     "execution_count": 294,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labels_pd.index = labels_pd['index']\n",
    "labels_pd.drop(\"index\", axis=1, inplace=True)\n",
    "counts_pd.index = counts_pd['Unnamed: 0']\n",
    "counts_pd.drop(\"Unnamed: 0\", axis=1, inplace=True)\n",
    "\n",
    "df = counts_pd.merge(labels_pd, left_index=True, right_index=True).dropna()\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "wUxSCyz7jBQf"
   },
   "source": [
    "+ **The cell-type are now One-hot encoded.**\n",
    "\n",
    "+ **labels and the counts are shuffled together.**\n",
    "\n",
    "+ **A split into train and test sets (80:20 split) is performed.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 295,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "XDTqBhcA7V8t",
    "outputId": "f30f081b-ccb4-4076-f774-f31c4f240951"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['CD14+ Monocyte' 'Dendritic' 'CD56+ NK' 'CD4+/CD25 T Reg' 'CD19+ B'\n",
      " 'CD8+ Cytotoxic T' 'CD4+/CD45RO+ Memory' 'CD8+/CD45RA+ Naive Cytotoxic'\n",
      " 'CD4+/CD45RA+/CD25- Naive T' 'CD34+']\n"
     ]
    }
   ],
   "source": [
    "categories = df['bulk_labels'].unique()\n",
    "print(categories)\n",
    "\n",
    "#one-hot encoding\n",
    "y = np.zeros((len(df), len(categories)))\n",
    "for i in range(len(df)):\n",
    "  cell_type = df.iloc[i]['bulk_labels']\n",
    "  pos = np.where(categories == cell_type)[0]\n",
    "  y[i, pos] = 1\n",
    "\n",
    "#remove label when processing input data\n",
    "X = df.drop('bulk_labels', axis=1).values\n",
    "\n",
    "#shufle and 80:20 split\n",
    "np.random.seed(100)\n",
    "permutation = np.random.permutation(len(X))\n",
    "X, y = X[permutation], y[permutation]\n",
    "\n",
    "X_train, y_train = X[:int(len(X)*0.8)], y[:int(len(y)*0.8)]\n",
    "X_test, y_test = X[int(len(X)*0.8):], y[int(len(y)*0.8):]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 296,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 854
    },
    "id": "l0Xrhu4LZNa9",
    "outputId": "0c62d1cb-7cfc-4e11-ec81-0a5b474f116f"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPkAAAP8CAYAAACEahi+AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8fJSN1AAAACXBIWXMAAC4jAAAuIwF4pT92AABtO0lEQVR4nO3deXhMZ/8/8PdMEtkjIqgIglDEVmurRWLfaVCqat/rUR6trbWVR3VVam3VrosilNr3Uvu+axJJtEWCWEISIp/fH36ZrzFLkskkM3N7v64r12XOOfc59+Cds92LRkQERKQsra0rQES5iyEnUhxDTqQ4hpxIcQw5keIYciLFMeREimPIiRTHkBMpjiEnUhxDTqQ4hpxIcQw5keIYciLFMeREimPIiRTHkBMpjiEnUhxDTqQ4hpxIcQw5keIYciLFMeREimPIiRTHkBMpjiEnUhxDTqQ4Z1tXwFHduXMHe/bs0X0uXrw4XF1dbVgjslepqam4evWq7nODBg3g6+ubZ8dnyC20Z88etG/f3tbVIAe0du1atGvXLs+Ox8t1IsU5VMiTkpIwbNgwBAQEwM3NDdWqVcPPP/9s62oR2TWHulwPDw/HkSNHMG3aNJQrVw4//vgj3n77baSnp6Nr1655WpfixYvrfa6C1+ABrzytg736bvdFi8r1Dy1v5ZrYh4dIwmkc0H1+/v9ObnOYkG/cuBHbtm3TBRsAwsLCEBsbiw8//BCdO3eGk5NTntXn+YdsHvCClyZ/nh3fnoW8bNkDSGX//kT/Y14/oHWYy/WIiAh4eXmhU6dOest79eqFf//9F4cOHbJRzYjsm8Ocyc+ePYsKFSrA2Vm/ylWqVNGtr1u3rkX7jo+PR0JCQrbKREZGWnQsorzmMCG/desWSpcubbDcz89Pt95Sc+bMwaRJkywuT2TPHOZyHQA0Go1F64heZA4T8oIFCxo9W9++fRvA/53RiUifw1yuV65cGT/99BPS0tL07svPnDkDAKhUqZLF+x48eLDBA73MREZGssUbOQSHCfmbb76J77//HqtXr0bnzp11y5csWYKAgADUqVPH4n0XLlwYhQsXtkY1ieyOw4S8RYsWaNKkCQYNGoR79+4hODgYP/30EzZv3ozly5fn6TtyIkfiMCEHgDVr1uCjjz7C+PHjcfv2bZQvXx4//fQTunTpYuuqEdkthwq5l5cXZsyYgRkzZti6KkQOw6FCTo6hWUC1PD3eln9PWlQur+tpKw7zCo2ILMOQEymOISdSHENOpDiGnEhxDDmR4hhyIsUx5ESKY8iJFMeQEymOISdSHENOpDiGnEhx7IVmI+w5ZT38OzGPZ3IixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1Ice6HZCHtOOa7s9iA8dykVVUJzpSpZwjM5keIYciLFMeREimPIiRTHkBMpjiEnUhxDTqQ4hpxIcQw5keIYciLFMeREimPIiRTHDioOhtMr2V52/y6T5C6Aq7lSl6zgmZxIcQw5keIYciLFMeREimPIiRTHkBMpjiEnUhxDTqQ4hpxIcQw5keIYciLFMeREimPIiRTHXmg2wt5klFd4JidSHENOpDiGnEhxDDmR4hhyIsUx5ESKY8iJFMeQEymOISdSHENOpDiGnEhxDDmR4hhyIsWxF5qNqNybjD3s7AvP5ESKY8iJFMeQEymOISdSHENOpDiGnEhxDDmR4hhyIsUx5ESKY8iJFMeQEymOISdSHENOpDj2QrMRlXtqOUIdXyQ8kxMpjiEnUhxDTqQ4hpxIcQw5keIYciLFMeREimPIiRTHkBMpjiEnUhxDTqQ4hpxIcQw5keLYC81G2FPLelTu0WcNPJMTKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHi2AvNwbDHlSGVv5s18ExOpDiGnEhxDDmR4hhyIsUx5ESKY8iJFMeQEymOISdSHENOpDiGnEhxDDmR4hhyIsUx5ESKYy80B5PXPa4s6fXGXmH2xeZn8vv372PkyJFo2rQpChUqBI1Gg4kTJxrd9vjx42jcuDG8vLzg6+uL8PBwREdH522FiRyMzUN+69YtfPfdd0hNTUX79u1Nbnfx4kWEhobi0aNHWLlyJRYuXIjLly+jXr16SEhIyLsKEzkYm1+ulyxZEomJidBoNLh58yYWLFhgdLvx48fD1dUVGzZsgI+PDwCgRo0aKFu2LL788kt89tlneVltIodh85BrNJpMt0lLS8OGDRvQvXt3XcCBp78gwsLCEBERkaOQx8fHZ/tqIDIy0uLjEeUlm4c8K6KiopCcnIwqVaoYrKtSpQq2bduGlJQUuLm5WbT/OXPmYNKkSTmtJpFdsvk9eVbcunULAODn52ewzs/PDyKCxMTEvK4WkUNwiJBnMHdpn5XLfqIXkUNcrhcsWBDA/53Rn3X79m1oNBr4+vpavP/BgwejU6dO2SoTGRlp9m0Akb1wiJCXKVMG7u7uOHPmjMG6M2fOIDg42OL7cQAoXLgwChcunJMqEtkth7hcd3Z2Rps2bbBmzRrcv39ftzwuLg67du1CeHi4DWtHZN/s4ky+adMmPHjwQBfg8+fPY9WqVQCAli1bwsPDA5MmTUKtWrXQunVrjB49GikpKRg/fjz8/f0xYsQIW1afyK7ZRcgHDRqE2NhY3edff/0Vv/76KwDgypUrCAoKQvny5bF7926MGjUKHTt2hLOzMxo2bIgvv/wShQoVslXVieyeXYQ8JiYmS9vVqFED27dvz93K2Lm8niaJnU0cn0PckxOR5RhyIsUx5ESKY8iJFMeQEymOISdSHENOpDiGnEhxDDmR4hhyIsUx5ESKY8iJFMeQEynOLnqhqeC73RcR8rJrlrdnrzDKKzyTEymOISdSHENOpDiGnEhxDDmR4hhyIsUx5ESKY8iJFMeQEymOISdSHENOpDiGnEhxDDmR4tgLzUr6h5aHlyZ/rh8nr+dCI8fHMzmR4hhyIsUx5ESKY8iJFMeQEymOISdSHENOpDiGnEhxDDmR4hhyIsUx5ESKY8iJFMeQEymOvdBshL3JKK/wTE6kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjLzQbcZTeZJb0lnOU7/ai4JmcSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx15oVvLd7osIedk1y9s7Sk8tR6knmcYzOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY690Kykf2h5eGny27oaVse50Bwfz+REimPIiRTHkBMpjiEnUhxDTqQ4hpxIcQw5keIYciLFMeREimPIiRTHkBMpjiEnUhw7qFgJp0kie8UzOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY690Kwku9MkWTL9EMBeYZR9PJMTKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHi2AvNRvK6Nxl7vb24bHom37lzJ3r37o3y5cvD09MTxYoVQ7t27XDs2DGDbY8fP47GjRvDy8sLvr6+CA8PR3R0tA1qTeRYbBryuXPnIiYmBu+//z42btyIGTNmID4+Hq+++ip27typ2+7ixYsIDQ3Fo0ePsHLlSixcuBCXL19GvXr1kJCQYMNvQGT/bHq5Pnv2bBQuXFhvWfPmzREcHIypU6eiYcOGAIDx48fD1dUVGzZsgI+PDwCgRo0aKFu2LL788kt89tlneV53Ikdh0zP58wEHAC8vL1SsWBFXr14FAKSlpWHDhg3o0KGDLuAAULJkSYSFhSEiIiLP6kvkiOzuwdvdu3dx/Phx3Vk8KioKycnJqFKlisG2VapUwbZt25CSkgI3NzeLjxkfH5/ty/7IyEiLj0eUl+wu5O+99x4ePHiAjz76CABw69YtAICfn5/Btn5+fhARJCYmomjRohYfc86cOZg0aZLF5YnsmV2FfNy4cVixYgW+/fZb1KhRQ2+dRqMxWc7cOqIXnd00hpk0aRKmTJmC//3vfxgyZIhuecGCBQH83xn9Wbdv34ZGo4Gvr29eVZPI4djFmXzSpEmYOHEiJk6ciLFjx+qtK1OmDNzd3XHmzBmDcmfOnEFwcHCO7scBYPDgwejUqVO2ykRGRqJ9+/Y5Oi5RXrB5yCdPnoyJEyfi448/xoQJEwzWOzs7o02bNlizZg0+//xzeHt7AwDi4uKwa9cuDB8+PMd1KFy4sNEn/UQqsGnIv/rqK4wfPx7NmzdHq1atcPDgQb31r776KoCnZ/patWqhdevWGD16NFJSUjB+/Hj4+/tjxIgRtqg6kcOwacjXr18PANi8eTM2b95ssF5EAADly5fH7t27MWrUKHTs2BHOzs5o2LAhvvzySxQqVChP60zkaGwa8t27d2d52xo1amD79u25VxkiRdnN03Uiyh0MOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIqz+fzklDeaBVSzdRVyzZZ/T1pUTuW/k2fxTE6kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjLzQyy5IeXnndu+tF6U1mKZ7JiRTHkBMpjiEnUhxDTqQ4hpxIcQw5keIYciLFMeREimPIiRTHkBMpjiEnUhxDTqQ4dlB5QXAqoRcXz+REimPIiRTHkBMpjiEnUhxDTqQ4hpxIcQw5keIYciLFMeREimPIiRTHkBMpjiEnUhxDTqQ49kJ7QeRlbzL2eLMvPJMTKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHi2AvNRlTuqeUIdXyR8ExOpDiGnEhxDDmR4hhyIsUx5ESKY8iJFMeQEymOISdSHENOpDiGnEhxDDmR4hhyIsUx5ESKYy80K/lu90WEvOya5e3ZU4vyCs/kRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOPZCs5L+oeXhpcmf5e1VnguN7AvP5ESKY8iJFMeQEymOISdSHENOpDiGnEhxDDmR4hhyIsUx5ESKY8iJFMeQEymOISdSHENOpDj2QrMSzoVG9sqmZ/KTJ0+iVatWKFGiBNzd3eHn54fXXnsNy5cvN9j2+PHjaNy4Mby8vODr64vw8HBER0fboNZEjsWmIb9z5w6KFy+OqVOnYuPGjVi6dCmCgoLw7rvvYsqUKbrtLl68iNDQUDx69AgrV67EwoULcfnyZdSrVw8JCQk2/AZE9k8jImLrSjzv1Vdfxb///ou4uDgAwFtvvYVdu3YhKioKPj4+AIDY2FiULVsWw4cPx2effZbndTx37hwqVaqk+3x6d3FerpNRSXIXB7FN9/ns2bMICQnJs+Pb5T25v78/4uPjAQBpaWnYsGEDunfvrgs4AJQsWRJhYWGIiIjIccjj4+OzfUUQGRmZo2MS5RW7CHl6ejrS09ORmJiIX3/9FVu2bMGsWbMAAFFRUUhOTkaVKlUMylWpUgXbtm1DSkoK3NzcLD7+nDlzMGnSJIvLE9kzuwj54MGDMX/+fABAvnz5MHPmTAwYMAAAcOvWLQCAn5+fQTk/Pz+ICBITE1G0aNG8qzCRA7GLkI8dOxZ9+/ZFfHw81q9fjyFDhuDBgwf44IMPdNtoNBqT5c2tI3rR2UXIS5QogRIlSgAAWrZsCQAYM2YMevTogYIFCwL4vzP6s27fvg2NRgNfX98cHX/w4MHo1KlTtspERkaiffv2OTouUV6wi5A/r3bt2pg3bx6io6NRo0YNuLu748yZMwbbnTlzBsHBwTm6HweAwoULo3DhwjnaB5G9sstmrbt27YJWq0Xp0qXh7OyMNm3aYM2aNbh//75um7i4OOzatQvh4eE2rCmR/bPpmbx///7w8fFB7dq1UaRIEdy8eRO//vorfvnlF3z44YcoVKgQAGDSpEmoVasWWrdujdGjRyMlJQXjx4+Hv78/RowYYcuvQGT3bBry1157DYsWLcKSJUtw584deHl5oWrVqli2bBm6deum2658+fLYvXs3Ro0ahY4dO8LZ2RkNGzbEl19+qftFQETG2TTkvXr1Qq9evbK0bY0aNbB9+/ZcrhGReuzynpyIrIchJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHi7HJkGEfUP7Q8vDT5bV2NF9KWf09aVO5FGfueZ3IixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1Ice6HZCHtOWQ//TszjmZxIcQw5keIYciLFMeREimPIiRTHkBMpjiEnUhxDTqQ4hpxIcQw5keIYciLFMeREimPIiRTHXmg24ig9pyzpLeco3+1FwTM5keIYciLFMeREimPIiRTHkBMpjiEnUhxDTqQ4hpxIcQw5keIYciLFMeREimPIiRTHkBMpjr3QbMRR5kJjjzLHxzM5keIYciLFMeREimPIiRTHkBMpjiEnUhxDTqQ4hpxIcQw5keIYciLFMeREimPIiRTHkBMpjr3QbMTS3l2O0nuN7AfP5ESKY8iJFMeQEymOISdSHENOpDiGnEhxDDmR4hhyIsUx5ESKY8iJFMeQEymOISdSHENOpDj2QrMR9iajvMIzOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIrjXGg2ovKcZpznzb7wTE6kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixbGDygsiLzuNsKOJfbG7M/mCBQug0Wjg5eVlsO748eNo3LgxvLy84Ovri/DwcERHR9uglkSOw65C/s8//+CDDz5AQECAwbqLFy8iNDQUjx49wsqVK7Fw4UJcvnwZ9erVQ0JCgg1qS+QY7CrkAwcORP369dGkSRODdePHj4erqys2bNiAli1bIjw8HL///jsSEhLw5Zdf2qC2RI7BbkK+fPly7NmzB3PmzDFYl5aWhg0bNqBDhw7w8fHRLS9ZsiTCwsIQERGRl1Ulcih28eAtPj4ew4YNw7Rp0xAYGGiwPioqCsnJyahSpYrBuipVqmDbtm1ISUmBm5ubxcfP7iV/ZGSkRcciymt2EfLBgwfj5ZdfxqBBg4yuv3XrFgDAz8/PYJ2fnx9EBImJiShatKhFx58zZw4mTZpkUVkie2fzkK9evRrr16/HiRMnoNFozG5rbn1mZYleVDYNeVJSEt577z385z//QUBAAO7cuQMAePToEQDgzp07cHFxQcGCBQH83xn9Wbdv34ZGo4Gvr29eVZvIodg05Ddv3sSNGzfw1Vdf4auvvjJYX6BAAbRr1w6rVq2Cu7s7zpw5Y7DNmTNnEBwcbPH9OPD0dqFTp07ZKhMZGYn27dtbfEyivGLTkL/00kvYtWuXwfJp06Zhz5492LRpE/z9/eHs7Iw2bdpgzZo1+Pzzz+Ht7Q0AiIuLw65duzB8+PAc1aNw4cIoXLhwjvZBZK9sGnI3NzeEhoYaLF+8eDGcnJz01k2aNAm1atVC69atMXr0aKSkpGD8+PHw9/fHiBEj8q7SRA7Gbt6TZ6Z8+fLYvXs3XFxc0LFjR/Ts2RPBwcHYu3cvChUqZOvqEdktmz9dN2bx4sVYvHixwfIaNWpg+/bteV8hIgdmlyF/EeT1VELsGfbicpjLdSKyDENOpDiGnEhxDDmR4hhyIsUx5ESKY8iJFMeQEymOISdSHENOpDiGnEhxDDmR4hhyIsWxF5qVfLf7IkJeds3y9uwVRnmFZ3IixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1Ice6FZSf/Q8vDS5M/14+T1HGrk+HgmJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixbEXmoPJ695klvR6Y483+8IzOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY690BxMXs+Fxh5ljo9nciLFMeREimPIiRTHkBMpjiEnUhxDTqQ4hpxIcQw5keIYciLFMeREimPIiRTHkBMpjiEnUhx7odlIXvcmoxcXz+REimPIiRTHkBMpjiEnUhxDTqQ4hpxIcQw5keIYciLFMeREimPIiRTHkBMpjiEnUhw7qFjJd7svIuRl1yxvz44mlFd4JidSHENOpDiGnEhxDDmR4hhyIsUx5ESKY8iJFMeQEymOISdSHENOpDiGnEhxDDmR4hhyIsWxF5qV9A8tDy9NfltXw+osmc6JPezsC8/kRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOPZCczCW9AoDLO8Zxh5ljo9nciLFMeREimPIiRTHkBMpjiEnUhxDTqQ4hpxIcQw5keIYciLFMeREimPIiRTHtusWSk1N1fv8EEmA5P5xz11KzXwjI5LkrpVrQln1EEl6n5//v5PbGHILXb16Ve/zaRzIk+NWCbW05NXMN6E8cfXqVVSvXj3PjsfLdSLFMeREitOISB7cSarnzp072LNnj+5z8eLF4erqCgCIjIxE+/btdevWrl2L4ODgTPfJctYrZ091TE1N1bu9a9CgAXx9fTPdp7XwntxCvr6+aNeuXZa2DQ4ORkhISLaPwXLWK2frOublPfjzeLlOpDiGnEhxDDmR4hhyIsUx5ESKY8iJFMeQEymOISdSHENOpDiGnEhxDDmR4th2PRcUKlQIEyZM0PvMcnlbzhHqmFfYC41IcbxcJ1IcQ06kOIacSHEMOZHiGHIixTHkRIpjyIkUx5ATKY4hJ1IcQ06kOIacSHEMOZHiGHIFPX782NZVIDvCkDuYVatWmV3/5MkTdO7cOY9qY9rSpUtzNA/34MGDcfr06WyVOXfuHE6dOmWw/PTp0zh//rzRMk5OToiPj7eojo6CIbei8+fPY+jQoWjRogWuXbsGAIiIiMCJEyesdox3330X+/fvN7pORNCjRw/s3r3basfLcPv2bd2fHzx4gIULF2L27NmIjIw0un2vXr1w9+5di493+PBhvPLKK3j11VexePFiJCcnZ1pm4MCBOHPmjMHys2fPYtCgQUbLvBA9rYWsYvfu3eLq6ioNGzaUfPnySVRUlIiITJs2TcLDw612nNGjR4ufn59cvHjRYN2AAQPE09NT/vzzT5PlNRqNaLVaoz8eHh5SrVo1mTNnjm77y5cvS3BwsGi1WgkJCZHLly9L2bJlRaPRiEajES8vL6PH02g0cuPGjRx91+PHj8uAAQPEx8dHfH19ZciQIXLmzBmT2/v4+Mjly5cNlv/111+SP39+o2WsUc8Mn376qSQmJlplX9bEkFvJ66+/LlOnThURES8vL13IDx06JIGBgZmW/+ijj/TClWHOnDkybtw4vWVdu3aVoKAguX79um7ZiBEjxNXVVbZt22b2ONOnTxc/Pz/p0qWLzJgxQ2bMmCFdunSRggULytSpU6VXr17i4uIi33//vYiIhIeHS/369eW3336Tzp07S5kyZaRFixZy7do1iY+PlzfffFOaNm1qcBxrhicpKUm+++47qVmzpmi1Wnn99ddl2bJl8vjxY73tvLy85OTJkwblT548KR4eHkb3rdFo5NixYxIbG2v2Jyu8vb11/+72hCG3kmf/gZ8NeUxMjLi6umZavmTJkkbPiAcPHpSgoCC9ZY8ePZKwsDCpXr26JCUlyaRJk8TZ2VnWrFmT6XHefvttmT59usHyb775Rt5++20REfniiy+kSpUqIiJSpEgROXTokIiI3Lp1SzQajfzxxx+6csePH5fChQsb7E+j0cjIkSNl0qRJZn+y6ujRo9KvXz9xdXWVwMBA8fb2lsDAQNm9e7dumwYNGsjQoUMNyg4ZMkTeeOMNo/s1d2Wj1Wp167Pi2X93e8Ix3qzEzc3N6D3opUuXsjTe1/Xr11G0aFGD5UWKFNHd32dwcXHB2rVr8cYbb+CVV15BdHQ0Fi1ahDfffDPT42zYsAGTJ082WN66dWuMGzcOANCmTRtMnDgRAJCQkIASJUoAAPz8/ODu7q5Xz6JFi+LmzZtGj/XTTz/BycnJZF00Gg3Gjx9vcn1SUhKWL1+O77//HqdOnUKzZs2wcuVKtG7dGklJSRgzZgx69eqF6OhoAMC4cePQrFkzREZGokmTJtBoNNiyZQu2bduGzZs3mzzOqlWr4OfnZ3J9dmg0Gqvsx5oYcitp1aoVPv30U/z8888Anv5j37hxAx999BHatm2bafkiRYrg1KlTCAoK0lt+8uRJ+Pv76z7v3btX9+dRo0ahT58+6Ny5M0qWLKm3rn79+kaP4+bmhgMHDqBMmTJ6yw8cOAA3NzcAQHp6Otzd3QE8fTCl1f7f81mtVqv3H9ncf+qjR4+icOHCJteb06dPH6xcuRLu7u7o3bs3Vq1ahVKlSunW+/j4YMyYMZg7d65uWaNGjbB582Z88skn+Oijj6DRaFCjRg1s2bIFDRs2NHms119/3aJ69u7dW+9zamoqRo4cCW9vb92yhQsXZnu/1saQW8kXX3yBsLAwFC9eHMnJyWjdujViYmJQpkwZTJkyJdPyHTp0wPDhwxEQEIBatWoBePqEecSIEejUqZNuu9DQUGg0Gr2nwj/99BN++ukn3WeNRoMnT54YPU7v3r0xaNAgXLx4EXXq1IFGo8GBAwfw7bff4r333gMA7Ny5E5UrV9aVWblyJXx8fAA8fUUXERGhuzox9QQ9p2e0y5cvY/78+ejYsSPy5ctndJuXXnoJu3bt0lvWuHFjNG7cOMvHyUk9xciTeXl6C2zxPnMDR2u1otTUVPz88884cuQI0tPTUbNmTbzzzjtwdXXNtOzDhw/RsmVL/PHHH/D29oZGo8G9e/fQoEEDbNiwAR4eHgCA2NjYLNWlZMmSRpenp6fjq6++wjfffKO7DShatCiGDx+O//73v9BqtYiPj4dWq4W/v7/eWdwUY79UtFotrl+/btEZ8tGjRxgyZAjGjBmjd/bODTmp5/O8vb1x6tQplC5d2go1sx6G3Er27t2LunXrwtlZ/+IoLS0Nf/75p8nL5+ft2LEDx48fBwDUrFkTYWFhVq9rhvv370NEdGdpa+rVqxdmzpypd+maHT4+Pjh9+rTB7cvzSpcujSNHjqBgwYIoVaqU2TNzxr17brHXkPNy3UrCwsJw7do1gzPC3bt3ERYWZvLy+XmNGjVCo0aNcqOKeo4fP46//voLrVu3BvC0gYurq6vBLylLLVq0KEflw8LCsGfPnkxD3qNHD93zgx49etjlgy9bY8itRESM/ge7e/eu7lL7eUuXLkXnzp3h6uqKpUuXmt1/9+7dARg++DJGo9EgLS3N6LqbN2/izTffxP79+6HRaPDXX3+hdOnSGDZsGLy8vDB9+nSz+84rb731FkaNGoWYmBjUrl0bnp6eeuszroyenbUk442ArXTr1i1XropyipfrOZTxhHXJkiV46623dGcV4OlDqmPHjsHf399oU9Nn7wfN3fs+e8+7ePFikyG/ePEiZs6cidTUVJMh79WrF+Li4rBo0SKEhIToLi83b96MESNG4Ny5c1n96rkqq38fz7pw4QIqVKhgtMzmzZvRvHlzq9XPkfBMnkNXrlwB8PRMHhcXp/ckOF++fKhXrx5GjhxptGx6errRP5vTs2dPg2W3b9/GhAkTMH/+fNSpU8fs2Xjr1q1Yv3697t13hnLlyiEuLi5LdcgLGX+v2VGzZk3Mnj1b7+/oyZMnGD16NGbMmIFHjx5ZsYbGmbqisykbNMBRUuvWreXOnTsWl9+zZ49BM00RkbS0NNmzZ4/RMo8fP5avvvpK/Pz8JDg4WH799ddMj+Pu7m60Zd6JEyfEx8fH4vrbg1mzZombm5t069ZNkpKSJCYmRurUqSNFixbNtLmviEh8fLx89dVXMnDgQElISBARkX379kl0dHSW6+Di4iLnz5+3+DvkBp7JrWTcuHHInz+/xeVNPbi7c+eO0Qd3a9aswciRI3Hr1i189NFHeP/99+Hi4pLpcWrUqIHNmzdj8ODBessXL16MOnXqWFx/Uz7++GMUK1bMoBfY3Llzce3aNXzyyScmy96+fRuzZs3CmTNnoNFoULlyZQwePBgFCxY0uv17772H119/HV26dEHVqlVx+/Zt1KxZEydOnECRIkXM1vP06dMICwtDgQIFEBsbiw8//BD+/v7YvHkzYmJisGzZMr3tTdX7yZMnmD17tu7f0VyLvjxj698yqtBoNFKtWjWZN2+e3L9/36Ly8fHxBsujoqLEy8tL9/no0aNSv359yZcvnwwZMkRu3ryZreNs375dPDw85MMPPxR3d3cZN26cvPnmm5IvXz7Zt29flvaRnd5W2WmT/6yTJ0+Kn5+fFC1aVNq1ayft2rWTokWLip+fn9FOKBlSU1Old+/eotFoxMXFRSIiIrJUzyZNmsjQoUMlPT1d7wrnjz/+MFpPjUYjgYGBEhQUpPej1WqlWLFiEhQUJKVKlcrSsXMbH7xZyblz5zB//nysWLECjx8/xttvv40BAwagevXqZstl98Gdk5MT3N3d8Z///MfkQybg/57GG7N792588sknOHz4sK7RzieffILQ0NAsfVcfHx+cPHkyS++D3dzccPHiRYNXYTExMShfvjxSUlKMlmvSpAny58+PFStW6BoTpaSk4J133sG9e/ewbds2gzKRkZHo1KkTEhMTsWzZMuzatQuTJ0/GkCFD8Pnnn5u90ilQoAAOHz6MsmXL6r3vjo2NRfny5Q36s/fo0QMXLlzAL7/8otdgx8XFBadOnULFihUz/bvJM7b+LaOa5ORkWbx4sbz++uui1WqlRo0a8v3330tSUpLR7UNDQyU0NFQ0Go3UrVtX9zk0NFSaNm0qAwcO1LsnzOjHbe4nq72mnnfhwoUsbZed3lYlSpSQtWvXGiyPiIiQYsWKmSzn6ekpp0+fNlh+8uRJ8fT0NFrG29tb2rVrJ7dv39Yt27NnjwQGBkqNGjXM1rNAgQJy9uxZERGDM3mRIkWMlvn+++/lpZdekpUrV+qWOTs7y7lz58weK68x5Lnou+++E1dXV9FoNJI/f34ZOXKkPHjwwOi2OX1wlxP//POP9OnTR1xcXLK0vZeXV5YfRg0fPlxKlSolhw8f1i07dOiQlC5dWoYNG2ayXP78+eXo0aMGyw8fPmzyAeHXX39tdPnNmzeldevWZuvZsWNHGTx4sIj83/dLTk6Wpk2byrvvvmuy3MmTJ6VcuXIycOBASUlJYchfBImJiTJjxgypVKmSuLi4yJtvvimbN2+WH374QcqWLSvNmze3Sb3u3bsn3bt3F39/fwkICJC5c+eKiMjUqVPFw8NDqlSpIsuXLzdatlevXno/Li4u0rFjR71lpjx48EAaNGggWq1W8ufPL76+vqLVaiUsLMzkLzwRkXbt2kn9+vX1nlPcuHFDGjRoIO3atTP7XVNTU+Xs2bNy7tw5SU1NNbtthujoaClatKjUqlVL8uXLJ02bNpXAwEApVqyYxMXFmS177949eeutt6RKlSp2GXI+XbeSffv24bvvvsPq1atRoEAB9OvXD/369UNAQIBum+rVq+PVV1/Vff7kk0/wwQcfwMPDw+xTZiDnT2nHjx+P7du3o3v37ti4cSPee+89/PHHH/jrr7+watUqtGjRwmRZyUFvKw8PD+zevTvbbfK//vprNGnSBCVLlkT58uUBPG3s89JLL+GHH34wWc9Jkybhiy++QEpKCkQE7u7u+OCDDzBhwgSzDWxKlSqFU6dOYc6cOShcuDDS09PRq1cvDBkyJNPOK97e3vjll18wZ84cbNu2zf5avdn0V4xCtFqtNGnSRNasWSNpaWlGt0lKSpKePXvqPgcFBemejpcsWdLgSW3GjzWe0pYqVUo2bdokIiKXLl0SjUZj9gxsTl6NgJKamioLFy6UESNGyH//+19ZuHChpKSkmNx+0qRJ4uvrK998842cOHFCTpw4IdOnTxdfX1+ZOHFirtfXXvHpupVERkYiODjY1tUwydXVFdHR0ShWrBgAwN3dHQcPHkTVqlWzva/MeltZ0ibfGoKCgvDpp5/i7bff1lv+448/YuzYsYiJiTFZtlSpUujRowd69OiR5e6tqampOH/+PF5++WV4eHjgwYMHWLFiBVJTUxEeHq77u7Y5W/+WUcXZs2eNvr89depUpvdoqamp0rRpU7l06VJuVc9gYMWcnI0zK/vssXLyFuDu3buydetWWbZsmSxZskTvx5h8+fJJZGSkwfLIyMhMx9mbMGGClClTRrRardSvX18WLVpk8o2IiMj58+elaNGiuvflkZGRUrlyZfHw8BB3d3fx9fXN1X/P7GDIreSNN96QZcuWGSxfsWKF1K9fP9Pyvr6+uXoJ/PzAiq6urvL+++9bNLDis80+c8v69eslf/782frlULZsWaMj3s6ePVvKli2bpePu2bNHevXqJT4+PuLl5SXdu3eXnTt3GmwXHh4urVq1klOnTsmQIUMkJCREmjZtKikpKZKcnCzNmzfXuzWzJYbcSiwZ8/tZ3bt3NzqKalZkpQWauXt+a977Z8eTJ09Mritbtqz06dNH/v333yzvb+bMmeLq6ir//e9/Zd26dbJu3ToZPny4uLm5yYwZM7JVt4cPH8ry5culatWq4uTkZLC+SJEiuld8d+/eFY1GIzt27NCt37Nnj920eGPIrcSSMb+f9c0330jBggWlZ8+eMmfOnCxdnmaw1/G+RZ6+fnv48KHB8ri4OJPDJItYfjvx3XffScmSJXVn/RIlSsj8+fOztY+MgDdp0kScnJykZMmSBtu4urrK1atXdZ89PT3lr7/+0n2OjY0VNze3bNc/NzDkVmLJmN/Pysm9a24/7U5ISJBPPvlEGjVqJOXLl5eQkBBp166d0duT55UpU0YqVKigN/PJunXrxM/PT8LCwkyWa9u2bZZ61Zly//59uXfvXrbK7N69W3ep7unpKe+++65s375d0tPTDbYNDAyUI0eO6D4/fwtz+vRpKViwoMX1tyaG3Eq2b98uTk5O0rJlS5k+fbp888030qJFC3F2dpbt27fn6rGz0wItu86dOyeFChXSdRbRaDTSsmVLqVWrlmi1WmnatKnZBif37t2Tzp07i4eHh8ydO1fef/99cXZ2lgkTJpi9XL927Zo0btxYvvzyS9m+fbvs2bNH78eYsLAwo7ctd+/eNfsLReTp7YyTk5PUr19fFi5cmGkno+bNm+saFBmzcOFCef31183uI68w5Fa0bds2qVevnnh4eIinp6fUr19f7z7NWnLSAi27mjZtKj179tS9+584caKuxdn58+elWLFiMn78+Ez3M3bsWF3PsK1bt2a6/R9//CEBAQHZurIxNTVTfHy8ODs7mz3ehAkTsvWLMiEhQW7dumVy/YYNG3L9l3tW8T25DT07GUJmnh3ttVevXnrrVqxYgbZt2+qNjJrTgRQzeHl54ejRo7pWZykpKfDx8UF8fDx8fX2xcuVKjB071uTspsDTcdv79euH2rVr49y5cwgKCsLPP/9sMDrNsypVqoSSJUvi448/RtGiRQ1GW3l2yOmMEW2CgoJw9OhRvckonjx5go0bN2LatGm4evWqRX8Hjo4ht7K9e/fqDXJQr149k9tmDMqY8U+Q8R/5+c8AzI72mpWhgOPi4lC8ePFsD01UpEgR7Nq1S9d1Mjk5GV5eXkhISICfnx+io6NRsWJFk11GBwwYgMWLF2PKlCn48MMPER8fj3fffRdHjhzB999/jw4dOhgt5+HhgVOnTqFs2bKZ1jGzwS01Gg0+/fRTfPjhh3rLrdGs+P79+9i2bRuio6Oh0WhQpkwZNG7cGF5eXpnWO8/Y8jJCJbdu3ZIGDRqIRqMRX19f3Tve+vXrmxzYISYmRvezfv16KVWqlMyaNUtOnjwpJ0+elFmzZknp0qXlt99+M3vsrDx402q1Fs0y2rZtW+natatuaKpRo0bpDaJw+PBhk10xRZ423T148KDB8k8//VTy5ctnslz9+vWzNGSTyNMHZrt27RKNRiNr1qyR3bt3637+/PNPk6/hnm1WbMmrxTVr1oifn5/B7UTBggWzPFhFXmDIraRHjx4SEhKi9xrtxIkTEhISIj169Mi0/Ouvv270P0ZERIS8+uqrZstmJeSWTiV87tw5KVy4sHh6ekr+/PnFzc1Nr55z5syRfv36mSxv7v39gQMHTK7btGmTVKlSRVatWiWRkZFZmkr46NGjJvsNWNuFCxfE1dVVwsPD5ciRI7pGMIcPH5YOHTqIq6trlvvn5zZerluJv78/1qxZYzBTyp49e9ChQweTM39mcHd3x+nTpw0uTy9fvoyqVasajEzyrEGDBmHy5Ml696LPy8l0QPHx8Vi/fj1SUlLQuHFjvPzyy9kqn92x2jLqm+HZS3H5/6Ohmrp9efz4MX766Se9Y3Xp0iVL499lx5AhQ3D58mVs3brV6PoWLVqgTJkymDVrllWPawmG3Erc3d1x4sQJ3QOqDBcvXsQrr7xiNqQAULZsWXTs2BGffvqp3vIxY8Zg1apV+Ouvv3JUP61Wi4iICBQoUMDsdlmdzimrTp06hYYNG8LV1RW1a9cG8HQix9TUVOzcudNkB5k9e/aY3W+DBg0MlsXExKBZs2aIi4vT/SK6dOkSihcvji1btmTa8eTJkyc4cuQIYmJiDIZvfr4jTZUqVTBlyhSTM9auX78eY8eOxZkzZ8weM0/Y9DpCIXXq1JEhQ4YYLB88eLDUqVMn0/K//PKLaLVaqVOnjvz3v/+VESNGyKuvvipOTk7yyy+/5Lh+Ga+erD1sVFJSksn31iIijRs3lg4dOuh1EU1OTpbw8HBp3LixRd/FlDfffFPq1asn169f1y27du2avPHGG/Lmm2+aLRsZGSnly5c3+HvSarVG/158fHzM3iJFRUXZzRDXDLmVbNu2TVxcXKR69eoyfPhwGT58uFSvXl3y5cuXpffCIiLHjx+Xbt26SfXq1eWVV16Rbt26ybFjx/S2iY6O1vtPvHnzZmnVqpVUqlRJWrdubfLdrEajkSNHjug97DP2k10nT540+8vBkrHaMmR3HPT8+fPrtULLcOjQoUz7D7Rt21ZatmwpN27cEC8vLzl//rzs3r1bqlevbnS02cweZF6/ft1om3dbYMit6Ny5c9K9e3epUaOGVK9eXbp37271oYBq164tGzZsEBGR1atXi0ajkdatW8uoUaOkffv24uzsLOvWrTMoZ+mDt8xkFnJLxmoTedpF18/PT8qUKSPOzs66s+bHH38s3bp1M1rG09NTr/nss/vK7BdKoUKFdL9Qvb29dd1EN23aJLVq1TLYXqvVGh1CO8P169ctHlDT2hhyO/Po0SO5evWqyafJzzZhrVOnjkyZMkWv/PTp042OTGppyEuVKmX2p1ixYmb/M1s6Vlt2x0EXEWnUqJF06tRJHj16pFuWmpoqHTt2lEaNGpn9nt7e3nLlyhURedrENWMM+itXroi7u7vB9s9eypv7sQcc483KHj9+jBs3bhjMbWaudRfwdO6v3r17448//tAbO02ee5qs0Wjw8OFDAE/n227Tpo3eftq2bYuPP/7YYP8lS5aEk5NTtr/PP//8gy5duphsaHPt2jV8//33JstbMlYbABw5cgSzZ882aORSvHhxXL9+3WiZzz//HI0aNUJQUJBuNphDhw7h4cOH2L59u9nvGRwcjMjISAQFBSEkJASLFy9GtWrVsGjRIqOzr1irRWFeYMitJKshNaVfv364ffs2lixZgmLFiplswVWrVi2sXbsWISEhqFChAo4dO4YqVaro1h85cgR+fn5G62eJkJAQ1K1bFwMGDDC6/tSpU2ZDXrp0aVy4cAErVqzAuXPnICL4z3/+g65du+omTTBGo9EYnaDw6tWrJqejql69Ov766y98++23umP17dsXQ4YMQaFChcx+z169euH8+fNo3LgxRo0ahZYtW2LhwoXQaDSYO3euwfbt2rXD0qVL0aNHD4P63LlzB0uXLjU6OaVN2PIyQiWNGjXSDWu8a9cuvVZXu3fvzrS8l5eXnDhxItPt9u7dK56envLRRx/J7NmzpWDBgjJ69GhZsmSJjB07VvLnzy8TJkzI+Rf6/3r16iUDBw40uf78+fNmpzsyNZHj48ePzT6Vt2Qc9NjYWKPdQjPWZUdcXJysXr3a6D2+yNOhrNu0aWOyfLt27eTzzz/P1jFzC0NuJVkNqSmlS5c2+hTamE2bNknp0qUNXoF5eHjIqFGjTP5Ht0RKSorZ8dEzY+op9M2bN83es1oyDrqlx7JE9erV5ffffze5/vfff8901pa8wst1KylcuLBF97wZxo8fjwkTJuCnn34yexkLAM2bN0dkZCSOHz+OqKgoPHnyBAEBAahZsyY8PT0troMxmdUlM2Jivu67d+/Cw8PDZDlLxkEXE+26kpOTTX6PrPYEfL6RUHR0NCpVqmRy+5CQEERFRWVp37mNLd6sZMmSJVi3bl2WQmpMWFgYTpw4AWdnZ5QvXx758uXTW79z505rVdUi2e1tld2JHHMiowfZpEmTMGLECL06PXnyBPv378edO3dw9OhRg7LP9wR8VsYvJ41Gg7S0NL113t7eOHHihMlhuE+ePIk33ngDSUlJFn8va+GZ3EoWL16MEydOoFixYhaFNCgoyGDmz6y6du0aYmNjERgYiMDAwCyXe/LkCRISEuDv7w9nZ9P/FSIiItC3b18kJibqLffz88OCBQvQvn17gzIZD/pEBHFxcXp/H/ny5UO9evUwcuRIg3IZfcMz8+zbiown3SKCn3/+We+KKl++fChdujTmz59vdD+mHkimp6dj4cKF+Prrr40+6CtXrhz27dtnMuQHDx5EuXLlsvRdcp0NbxWU0rNnT7M/1jJp0iTdPfLjx4+le/fuumaYWq1WOnfurPeeOMOSJUv0mpaOHz9evL29RavViqenp4waNcro8XLa2yq7Ezlm9t7ZXPPbSpUqmR2tJat27Ngh1apVE09PT/n444+Njr8+bdo0KVGihNGmrdHR0VKkSBGZNm1ajutiDQy5nbl165YcPHjQ5HRAzz5c+vrrryV//vzyww8/yNmzZ2XJkiVSsGBB+eKLL8yWW7Jkibi6usq4ceNkw4YNMmnSJHF3d5cFCxYYlHvvvfekSZMmJuvbvHlzee+997L03ZKSksxOWCDytJFJqVKlZOrUqQZvKDJ7W5Gd4ZuNuXTpkrRp00a0Wq288847eqOxPi8lJUVq1aol7u7u0q9fP5k+fbpMnz5d+vbtK+7u7lKzZk2zUzrlJYbcyiyZUVPk6eyfXbt21Z2pMs4Q/fr105v04NmWazVr1pRvv/1Wbz8LFy6UypUrG+z/2XKvvfZallvKVa5c2Wgz2Qy//fabVKpUyex3W7RokZQqVUp3Ni5VqpQsXLjQ6LbHjh2Tfv36iZeXl1SpUkXmzJkjd+/eNbv/DM7OzrpZZLPj1q1bMmTIEHFxcZE33njDaPt3Y5KSkmTYsGHi7++ve8Ph7+8vw4YNy3QgyLzEkFvJkydPZMKECeLh4aG7rPTw8JDx48ebHZU0w/Dhw6Vq1aq69+AZIV+7dq288soruu2eDau/v7+cOnVKbz+RkZHi5eVlsP9nyxUuXNig48ulS5eMduLIaW+r2bNni6urqwwbNkwiIiIkIiJC3n//fXFzc5NZs2aZLHfv3j2ZPXu2VK1aVby8vKRPnz6Zhm/9+vXStm1bcXZ2lqCgIJk6dapcu3bNbJmvv/5a/Pz8JDg4WFatWmV2W3Pi4+NzpW+ANTDkVpLTGTWDgoLkjz/+EBH9kV4uXbqkFyKNRiNr166VPXv2SIkSJeTw4cN6+zl37px4e3sb7F+j0cixY8ckNjZWSpUqZTTkxjpx5LS3Vbly5WT27NkGy2fNmiXlypUzWe5ZBw4ckIYNG4pWq5Xbt29nuv0///wjEydOlJIlS0q+fPmkY8eOJoeS0mg04unpKe+8847BKLi5MfqtLTDkVlKyZEn58ccfDZavWLHC6Awcz3Nzc9N1kHg25GfPntUL3/P9nb/++mu9/Sxfvtzk5fqzD6+en1UkIiJCypcvb1Aup72tcjIJ4ZMnTyQiIkI3fn3z5s0lOTnZbJlnpaeny5w5c8TV1VW0Wq2UKVNG5syZo3dl1aBBAwkNDc30x5HxFZqVXLt2TTfyybPq1KljskPFsypUqID9+/cbvEaLiIjQGz3l+Vc+z7+nfvLkicGopACwa9cuvc9FixbV+xwbG4s+ffoYlBMRvPTSS5nW35SiRYviyJEjKFOmjN7yw4cPG+34ATx9jbZgwQL88MMPSE9PR+/evTFnzpwsv2JMT0/Hhg0b8P3332PTpk2oUqUKBgwYgH/++Qfjxo3Dn3/+iWXLlgGAVd7T2zuG3EpKliyJrVu3YtCgQXrLt2zZkmkPNAAYPXo0Bg0ahLt37yI9PR3bt2/HxYsXMXv2bKxatUrvOOaYmu/b2HBJz3r//feNLs9pb6sePXpg8ODBSEhI0NVh9+7dmDhxIoYMGWKwfevWrbF161aEhoZi5syZaNeundl3+M/K+OWwaNEi3Lp1C507d8aff/6p98u3WbNmaNq0qS7kLwK2eLOSb7/9Fh9++CHee+89vf/Mc+fOxWeffYahQ4dmuo9ly5ZhwoQJiImJAfC0W+X//vc/dOvWzWBbhxjvG0+vLMaPH4+vv/4ajx49gojA1dUVw4cPx+TJkw2aAmu1WpQoUcLsGPKA8cZFzs7OCA4OxoABA9CzZ0+j49ndu3cP7dq1M7iyURlDbkXff/89/ve//+labRUvXhwfffQR+vfvn6Xyt27dQnR0tO4/uqnukZa0QAOAX3/9FStXroS/vz8GDhyodxtw8+ZN1K5dG9HR0XplMrpNZtal0sfHx+x3S0lJQWRkJEQEZcuWhZubm9HtevbsmaUJIIxdYezYsQONGjXKtOwLx4bPA5SV3Rk1//rrL2nUqJE4OTnpHo45OTlJ06ZNDV5fWdoCbcWKFaLVaqVly5ZSt25dcXV11ZsS2dQDtKx0qfzss89Mru/Vq5fRv4ukpCSrPbV++PChrFu3zuj79Lt378q6devspmGKLTDkNnb79m0pWrSolCpVSr744gvZvHmzbNq0ST7//HMpXbq0FCtWTO+1kaUt0GrUqKHXEm7x4sXi5uYmS5cuFRHTIa9Ro0aOulSaegWXkJBgte6fs2bNknr16plcX79+fZk3b55VjuWIGPIcCgsLy9KPKZMmTZKXX37Z6Nnu7t27UqFCBZk8ebJumaUt0Ly9vQ2uCtauXStubm6ycuVKkyH39fU1O+BCTEyM+Pr6mlyv0WgMXsGlp6dLRESE2emVsuPVV1+VlStXmly/atUqee2117K8v08//dTszC+ORpv5BT2Zs3v3bsTExKBkyZJmf0zZvHkzRo0apTcjaQYfHx98+OGH2Lhxo25ZbGxspv2YjfXkcnNzM7iHb9euHRYtWoQePXrgt99+M7q/tLQ0o8MwZUhMTMTjx48Nlmu1Wjg5OUGj0eCll16Ck5OT7sfZ2RkdOnRA165dTe43O6KiolCzZk2T6zOGhcqqqVOn4vbt29aoml3gK7QcGjx4MFasWIEDBw6gf//+6Nmzp9Ex1ky5dOkSXn31VZPrX331VXzwwQe6z0lJSWafoHt6euLBgwcGyytVqoR9+/ahRo0aesu7dOmC+/fvY/DgwUb3Z2mXykWLFkFE0Lt3b3zzzTd6D+0yun9mDLaYU0lJSWYf1j18+FA3+GVWiGrPom19KaGChw8fyg8//CC1a9cWd3d36datm25I38w4OzubbV997do1cXFx0X22tAXaokWLTI5XLiLy1VdfGR2rLaddKletWmW066s1hYSEyK+//mpy/bJly6RixYpZ3l9WJpB0JAy5lR0/flwGDBggPj4+EhISkun4aNkNbV6P953TLpWDBg0y6ERjbWPHjpWQkBCj/dYTExOldOnSMnbs2Czvb8WKFZl2iXUkfE9uZWlpaVi9ejW++eYbHD16FDdv3jQ5hDDw9N61bt26BiPJZHj06BEOHDigG9J5yZIlWapHjx49zK6/efMmNBqN2ZlFMzx48AAff/wxli9fjlu3bgEAChYsiG7dumHy5Mlmbx9q1qyJEydOoFatWhg4cCA6d+6sNxRUVlSuXBkbN25E8eLFja6/e/cuqlevjqSkJLz//vuoWLEiAODcuXP49ttv4enpiWPHjsHX1zdbx1UFQ24lUVFR+O6777Bo0SK4u7ujT58+6Nu3LwICAsyW69WrV5b2b43B/G/evInRo0dj9erVuHfvHgAgf/786NSpE6ZOnZqlwCckJEBEsjUF8okTJzB//nz89NNP0Gq16NatGwYMGGD2AeKzvL29cerUKbOt4K5fv45BgwZh/fr1uokttFot2rRpg9mzZ2f675AhNTUVERERiImJQYkSJdC+fXuzA046BJteRyjg119/lcaNG4uLi4u0bt1a1q9fn6X+45ZKTEyUGTNmmLw0nTFjhtFGIcnJyVKhQgXx9fWV9957T+bNmydz586VwYMHi6+vr4SEhGSrh5clkpKS5LvvvpOaNWuKVquV119/XZYtW2Z0XPZnZeceOTExUQ4fPiyHDh3K0muwsLAw3Xa3bt3SzWyaMRBEmTJl7LafeFbxTJ5DWq0WQUFB6N27t9mOKKY6jmTXp59+igMHDph85dW+fXvUrVvXYJDE6dOn48svv8TBgwcNLnv//vtvvPbaaxgxYgSGDRtmlXqacuzYMcyfPx9Lly5FoUKFcPfuXeTPnx/Lly832YnG29sbp0+fznR+cUtotVpcv34dhQsXxtChQ7F161Zs3LgRpUuXRmxsLNq2bYt69eph1qxZVj92nrH1bxlHZ26+75zM+22KpS3QGjRoYDBU1LO+/fZbadCggTWqaOD+/fsyd+5cqV69ujg5OUnLli1l3bp18uTJE7l7964MHjxYSpUqpds+Kw8XreXZEXMqVqxoMDrMxo0bJTg42GrHswW+J8+h5yc2zG1RUVEWDep/4cIFgwkCntWgQQPd+OXW1KdPH6xcuRLu7u7o3bs3Vq1apXdG9vHxwZgxY/TmG3u2h5iIoGXLlliwYAGKFStm9foB/ze++o0bN1ChQgW9dRUrVsTff/+dK8fNKwy5g7G0BVpiYqLZSf/8/f1x9+5dq9TxWZcvX8b8+fPRsWNHk28QXnrpJb1gP3/Z7uTkhFdffTXT7qeW+vLLL+Hp6QkXFxckJCTorbt9+7bVZ6XJawy5g7G0BVpaWprZwRe0Wq3BLCHW8Mcff2S6jbOzs9lBLbLS9dRSJUqUwMqVKwE8bfp74sQJvbrs2rULL7/8cq4dPy8w5A7mrbfewoQJE1C/fn2DM9uVK1cwceJEDB8+3GjZoUOHmnxHnZycbNV6njhxAsOGDcNvv/1mtB96u3btMGvWLFSuXDnTfUk2ng2npqbi999/x19//YUBAwbA19cXUVFRKFCggNHmxhkDdJhSt25dhIWFZfn49ohP1x1Mamoq6tWrh7Nnz6Jbt256DT9WrFiBkJAQ7Nu3z2A+ttDQ0CydEa01Ykr37t1RokQJTJkyxej6CRMmIC4uLkvv//ft24datWplOsdcTEwMmjRpgn///Repqam4fPkySpcujeHDh+Phw4cmp0pSnm2f+5ElHGFQ/7Jly5odJ/3o0aNWf2r95ptvyltvvSUpKSl679Z37twpZcuWNVv2/v37Rts3PHr0yOw86o6AIXdw2RnU/+7du0b/I2e8yrImNzc3+fvvv02uj42NFTc3N7P7OHPmjKxatUpWrVolZ86cyfSYhQoVknPnzomIfgOaK1euiLu7u9Eyd+7ckZYtW4qTk5N4eXnJhx9+qNc4J7Mhpx0B+5PngmnTpuHOnTt5cqxChQplqYnpunXr8Morrxi9905OTkb16tX1+q3nlLe3t9mhqK9cuWK0Dz0AnD17FjVq1EDVqlXRqVMndOrUCVWrVkXNmjVx7tw5k/tMSUkx+nAxISHB5LOIiRMn4tKlS/jpp5/w1Vdf4ccff0S7du303mCIo9/R2vq3jIqMjcJia82bNzfbGGbOnDnSsmVLqx2vVatW8sEHH5hcP2DAAGnVqpXB8piYGPHz85MKFSrIkiVL5Pjx43Ls2DFZsmSJVKxYUQoWLCgxMTFG99m8eXPdvHFeXl4SHR0tIiJvv/22tG/f3miZUqVKydatW3Wf//33X6lSpYq0adNGHj9+rMSZnCHPBfbYH7l48eJy/vx5k+svXLgggYGBVjve1q1bxdnZWf73v//p9Sd/9OiRTJ06VZycnPTClaFPnz4SGhpqtD17WlqaNGzYUPr27Wv0mMePHxdvb2/p1KmTuLq6Sq9evaRGjRri4+MjZ8+eNVrm2ZlrMty4cUMqVKggnTp1kr///pshJ0PPnkXshaurq9mpeKOiojKdtii7pk2bppv/vGrVqlK1alXx9PQUJycnmTp1qtEygYGBJqcmFhHZtWuX2V9GFy5ckF69eklISIhUqFBB3n33XbPzpwcFBcmOHTsMll+9elVKlSolLVq0cPiQ8xWaFfTu3Vvv8/Lly9GuXTu9e86FCxfmdbX0lCxZEosXLzb5znfjxo0YNGgQYmNjrXrckydPYunSpbh06RJEBBUqVMC7776LatWqGd3e1dUVUVFRCAwMNLr+77//RpkyZZCammqV+nXv3h3e3t6YPXu2wbqoqCi88cYbiI+P1/Xnd0RsDGMFxn5PytOrJBvUxrhGjRrh66+/NhnyL774Ag0bNrT6catVq2Yy0Mb4+fnh77//NhtyYzOjAE+bw/bq1QudOnXKclPUkSNH4vjx40bXlSlTBjt37tS1iHNYtryMUJU93pNfvnxZvLy8pGnTprJ//35JTEyUxMRE2bdvnzRr1kw8PT3l4sWLtq6mdOnSRTp27Ghy/VtvvSWdO3c2uq5Hjx7i7e0tXl5e0r17d9m1a1cu1dKxMOS5wB5DLvK0UUhAQIBBt82iRYvK9u3bbV09ERE5ffq0uLu7S7NmzWTfvn26X0Z//PGHNG/eXNzd3eX06dMmyz948ECWLFmim888KChIxo8fn+m/x71792T16tXyxRdfyJdffikRERF207AopxjyXGCvIRd5OkLMunXr5PPPP5fPPvtMfvvtt1wfESa7tmzZIkWLFjX4ZVSkSBHZtGlTlvcTFxcnU6ZMkTJlyoiTk5PJ7dasWSN+fn4G4wAULFhQIiIirPCNbIv35LmgW7dumU4AmNcmTpyIDh06oHLlymjbtq2tq2NW06ZNER0djS1btuDy5csQEZQvXx5NmzY1OVGiMXFxcYiJicHNmzdNlrt48SLefvtttGrVCmPGjEHlypUhIjhz5gw+++wzdOnSBSdPnkT58uWt9fXynq1/y1DeaNGihbi6ukpwcLCMHDlSDh8+bOsqWczcmHBXrlyRSZMmSXBwsGg0GqlXr54sWLDA5ASUls4t50gY8lySnp5u6yoYuHfvnqxYsUI6dOggnp6eEhgYKEOHDpU9e/bkSX337duX49lFly9fLqVLlza6rn79+uLk5JTl+3ARy+eWcyQMeS5xcXEx28LM1pKTk2X16tXSrVs38fX1lcKFC0u/fv1k8+bNmY6eaqmsNvdduHChdOzYUbp27arrybZv3z6pWrWquLq6mmzx1qNHj2w/Uffx8TFbp6ioKPHx8cnWPu0NG8PkkKlx0SZNmoRBgwbpOo+MHz8+L6uVLY8fP8b27dsRERGBdevWYciQIRg3bpzVj5OV8dNnzpyJDz74AJUrV8aFCxfg7OyML774AqNGjcKAAQMwbNgwFC1a1Gp1cnJywrVr10x28rlx4waKFSuWK6Pm5BWGPIe0Wi2KFStm0PspLi4ORYsWhYuLCzQaDaKjo21Uw+xJT0/H7du34e/vb/V9ZyXklStXxpAhQzBgwABs3boVzZs3x6uvvoo1a9bgpZdeMth+6dKl6Ny5M1xdXbF06VKzxzc2LLaTkxOuX79ucvy7GzduICAgwKFbvDHkOdSjRw9cuHABv/zyi94opC4uLjh16pRu5BZ7tH//ftSsWTPTEVcs9XzoBgwYgMmTJ+udNZ8Pnre3N86cOYOgoCCICFxdXbFjxw7Uq1fP6DGeHTddqzXdc1qj0RgNqlarzdKIOQz5C27BggUYN24cZs6ciU6dOgFwjJD7+Pjg5MmTuTYK6vOTIcTFxSEgIEB31WPsCufZ0AJZO/vnhLXmlrNnfE9uBX379kWtWrXw1ltvYefOnfjmm29sXaUsye3f71euXNH77O3tjT179mQa2AMHDujap6enp+Pw4cMGY5+bG0M+Oxw5vFnFkFtJ1apVcfToUfTt2xe1a9e2dXUcWnh4uN4voK5du+qtN3XpneH69euIiYkxGJ/e2C+GO3fuYOnSpejRo4fRUWWXLl2Knj172l3jpuxgyK3I29sbv/zyC+bMmYNt27bZ3X+M5++R09LSsGbNGrP3yHnt+bN/dsTHx6Nr164mR5w19oth7ty5OHDgAIYOHWqwztfXFzt37kRKSorB3HIOxUav7sgGgoKC9H60Wq0EBgbqPj87H1luCAkJkbi4uFzbf7du3aROnTpy6NAh8fT0lC1btsjixYulbNmyJtu8Wzq3nCNhyHNZUlKS3Q7pa88daZ6VlpYm165dy7SRTkBAgOzbt09Enja8uXz5soiIrFy5UsLCwoyW8fX1ldjYWJP7jImJEV9fXwtrbh84Wmsui4yMdPgZOPLS0qVL9UZ9mTBhAgoUKIBixYrB19cXo0ePNln27t27ummZfXx8kJiYCACoVasWDh06ZLSMpXPLORKGnHJNbGwsIiIicOPGDQBPR3X56KOPMGbMGJOh69Wrl27ixaVLl+Kzzz7TTbc0cuRIzJw5Ez/88IPRskFBQYiLiwMAlC1bFuvWrQMAbNmyBb6+vkbLZMwtZ4qpueUciq0vJRxdqVKlzP4UK1bMbgcCzM175B07doibm5toNBrx8/OT48ePS0BAgAQHB0u5cuXExcXF6L3ws/OFv/baazJlyhS99dOnTzd5j/zxxx/rtl+7dq04OTmJv7+/aLVa3VDNz5s2bZqUKFHC6G1LdHS0FClSRKZNm5at725v2Bgmh1xdXdGlSxeT736vXbuG77//3qFbTFmifv36CAkJweeff445c+Zg1qxZaNGiBb777jsAwPvvv49jx44ZnEWfbQxTpEgRbNq0CdWrV9etv3z5MmrXrp2lySsOHTqE/fv3o3z58mjZsqXRbSydW86h2Pq3jKN75ZVXZN68eSbXnzx50m7P5Lkpf/78cunSJRERSU1NFScnJzl69Khu/YULF6RAgQIG5TQajRw7dkxiY2OlVKlScuzYMb31ly5dEk9PT6vW1RHmlssJvifPoWrVquHkyZMm1+fLlw8lSpTIuwplU8OGDbFo0SKULFnSqvt9tj24sbbh5tqL16pVC8DTFnlHjx7VO5OfP39e93AtQ8Z9eGZM/Tt4enpi+vTpmD59OhISEiAiWZp6ylEw5Dk0d+5cs5fiFSpUyFEDD2vZu3evyeV79uxBUFAQAOs1F61atSpmzJihu1wPDAzEggULUKNGDQDA/PnzUaFCBYNyzzdkeb5baWxsLPr06aO3LCgoKNNOJhqNxmh30aCgIISHh6Njx46oW7euyd5ojoz35C+IjN5W5v65M2sumh27d+9Gq1atkJKSAjc3N+zatQt9+vTBo0ePoNVqERkZidWrV1tlvLk9e/aYXLd9+3Z8/fXX0Gg0SEpKMli/cuVKREREYOPGjfD09ET79u0RHh6OsLAwODk55bhu9oAht5L79+9j27ZtiI6OhkajQZkyZdC4cWN4eXnZumoAgIoVK6JcuXL44osvkC9fPgBPL4fLli2LTZs2oWzZsgBg1cv2uLg4HD16FDVr1kSJEiVw8+ZNzJw5E8nJyWjbtq3J7qPGLF68GG+++aZB+3JTLl++jA8++AC///47unbtik8//dTkhA0A8OjRI2zZsgURERH47bffAABt2rRBeHg4mjVrpvs7c0i2exygDkcY0vfBgwfy7rvvSo0aNfReFzk7O+vm9LZnWR1O69atW/Kf//xH8uXLJ2+88YZu+KjsSEtLk23btsmgQYMkICBAfHx8pHPnzhbtyx4w5Dl04cIFcXV1lfDwcDly5IikpKRIcnKyHD58WDp06CCurq5mJ9zLa/Pnz5ciRYrITz/9JCK5H/LnJy1Yu3at2SfWptobaDQaCQwM1H1+3uPHj+Wrr76SAgUKSJkyZeTXX3+12nfYt2+f/Pe//5UFCxZYbZ95iZfrOTRkyBBcvnwZW7duNbq+RYsWKFOmDGbNmpXHNTPt5MmT6NSpE0JDQ7FkyRKcPHkyVwa3iIiIQN++fXXNSzP4+flhwYIFaN++vUEZZ2dnNGvWTK+7rohg8uTJGDhwoO6p94QJE/SOM3LkSNy6dQsfffQRhg4dChcXF6t/H0fFkOdQlSpVMGXKFJMPkNavX4+xY8fizJkzeVwz8+7du4e+ffti27ZtOHjwIF5++WWr7v/ixYuoVq2ayUkLNmzYYHTSgh07dqBnz54YNmwYRowYoVtubqQdrVYLDw8P9O7d2+zYdFkdTHPw4MH45JNPcmWcO1tgyHMof/78OHHihMkWb9HR0XjllVd07bFfFDm5wvn333/RtWtXeHl5YenSpfDz8zMb8qy+QsvqYJq5PSxWXuN78hxKSkoy+wTd09MTDx48yMMamff8W4Dg4GA0atTI6m8B9u7diylTpphcP3jwYIwdO9bouoCAAOzcuRNjxoxBtWrVsHz5crMhjomJyWl19ah23mPIrSCzs4i9/Kex5B7ZUrGxsahUqZLJ9SEhIWZbqmm1Wnz22WeoX78+OnXqlOdt/7Mygquj4OV6DjnKkL6W3iNbypqTFvzzzz84evQoGjduDE9PT6vU71lhYWF6/4Z79+5FrVq14O7urlu2c+dOqx83r/BMnkOLFi2ydRWyZNasWahfvz5Wr16tt7xWrVpYtWoVWrRogVmzZln1LYA1rnAyZiRt166dtaplIDQ0VO/z/v37Ubt2bfj5+eXaMfOULd7bUd7L64n9NBqNwfzixn6MSUhIkD59+oivr69uuwIFCkj//v3l5s2bVqujKY4yLFZW8UyeQ44ypG9O75Gzy9IrnJSUFNSvXx/Xrl3DO++8o3db8eOPP2L//v04evRotuYpf+HZ+reMo5s6daq0adPG5Pp27drJZ599loc1Mk6r1epGXDHm+vXr4uTklIc1Mu7rr7+WgIAAoyPWXL16VQIDA2X69Om5WgfVzuQc4y2HVq9ejYEDB5pc379/f6xcuTIPa2RaXr4FuHPnDmbOnGm0fUDGunv37hmsW7duHcaMGWPQZxwAAgMDMWrUKKxduzbT4w8ePBg3b960qO5jx45V534c4Jk8pxxlSN+c3CNbwtIrnMKFC8upU6dMljt9+rQUKlQo0+NndS70FwHvyXPIUYb0zeu3AKtXrzY5dzvw9Apn/PjxBjOTJCYmmh24wd/fP0utB8VKVyV//PEH5s2bhxUrVlhlf7bAkOdQxpC+wcHBRtfby5C+eT2xX1RUVKYP+qKiogyWp6WlGcz1/iytVpuld+uA5Q1a7t27hyVLlmD+/Pk4f/48qlWrZtF+7AVDnkNvvfUWJkyYgPr16xu0db5y5QomTpyI4cOH26h2/yev3wLk5Apn6NCheg1RnpWcnGx0+fMNWpKTk9G1a9dsNWg5cuQI5s2bh59//hkpKSl4//33sXr1aqt33slrDHkODRs2DKtXr0alSpVMDuk7bNgw21YSeT+xn6VXOPXr18f169fN7tvYOHSWNmh5+PAhVqxYgblz5+Ly5cvo0KED1q9fj2bNmqFv374OH3AAfPBmDY4wpG9eT+xn60kLsvoaLH/+/PLGG2/IggUL9P6tHGXEnKzgmdwKHGFIX0vvkS2Vkyuce/fuwcvLC1qt/hve9PR0JCUlWbVh0ePHj+Hu7g4PDw91B5qw9W8ZyhteXl7y119/mVx/4sQJu5i0YO3atVK6dGlJSkoyur8yZcqYvSLJkNUzeWJiokyfPl0qVKggvr6+MnDgQDl48KC4uLgocyZnyF8Q1atXl0WLFplcP3fuXHnllVdy7fjx8fFmW9xlaN68uXz77bcm18+ZM0datmyZ6X6mTp0qiYmJ2ami7Ny5Uzp37iyurq6i0Whk1KhREh0dna192COG/AVh63vkrCpevLjZUVkvXLgggYGBuVqH+Ph4mTp1qpQuXVq0Wq3UqVMnV4+X29if/AXhKBP7ubm5ITIy0uQY6dHR0ahYsSJSUlKyvE9LG7SICDZv3ozZs2djw4YN2SprV2z8S4bykCO8BShRooTs3LnT5Prff/9dSpQokel+7t69KzNnzpSQkBDRaDQW34osW7ZMSpcubVFZe8GQv6Cyeo+c13r16iWtW7c2uT40NFR69uxpcv3hw4eld+/e4uHhIVqtVoYPHy4XL140e8yFCxdKx44dpWvXrrqZV/ft2ydVq1YVV1dX6devn2Vfxk4w5GRXLl++LF5eXtK0aVPZv3+/JCYmSmJiouzbt0+aNWsmnp6eBqF98OCBfPfdd/LKK6+Ip6endO/eXXbs2JGld90zZswQFxcXqV69uri7u4u3t7fMmzdP8ufPLyNHjpR///03N79unmDIye7s3LlTAgICDHrIFS1aVLZv326wfU4atFSqVEk3v/yWLVtEo9HIa6+9JteuXbPul7IhNoYhuxMWFoaoqChs3boVly5dgoigQoUKaNKkidERYXLSoCUmJgbNmjUDADRp0gTOzs747LPP8NJLL1nlu9gDhpzsysSJE9GhQwdUrlw5y9Ma//PPP1i8eDEmT56MwYMHo0uXLujZs2eWeqE9ePAAHh4eAJ72WnN1dUWxYsVy9B3sDV+hkV1p2bIldu7cieLFiyM8PBwdO3ZErVq1slx+165dmD9/PtauXYtHjx5h5MiRGDBgAEqVKmV0e61Wi4iICBQoUADA05ldfvjhBwQEBOhtZ6xTjKNgyMnu3L9/H+vXr8eaNWuwefNmFChQAOHh4ejQoQPq1auXpTN0QkICFixYgAULFiAmJga1atXCwYMHDbbLGDffXAw0Go3Nx83PCYac7FpKSgo2btyIiIgIbNiwAfny5UO7du3QoUMHNGrUyOwAE0DmDVpiY2OzVI+SJUtaVH97wJCTw3j8+DG2b9+OiIgIrFu3DkOGDMG4ceMyLbd8+XJMmDDBqr3sHAlDTg4pPT0dt2/f1k0vvGjRImzcuBH58uXDf//7X9SoUQP79+/He++9h4sXL6J79+747rvvsrTvJ0+eICEhAf7+/pleKTgEm7y4I8qGffv2SUpKisn1OWnQsmTJEr19jx8/Xry9vUWr1Yqnp6eMGjXKqt/FFhhysnuZDa+ckwYtz046sWTJEnF1dZVx48bJhg0bZNKkSeLu7i4LFiywzhexEYac7F5mA0B4eXnJlStXREQkPT1dXFxcZO/evVnat0aj0YX8tddekylTpuitnz59ulWHxbIFzqBCDs9aDVqioqLQokULvWUtW7ZEZGSkVeppKwo8VSDVLF26VO9zWloa1qxZozduXvfu3fW2OXDggK5BS3p6Og4fPoy///5bbxtTDVr+/vtvpKSkmJz7PKvjvNsrPl0nu/N867S4uDgEBATonnRrNBpER0fr1uekQUtGWeDpO/V58+ahf//+uvVr167FmDFjcOHChRx9J1vimZzszpUrV/Q+e3t7Y8+ePQaTV5jaPjt27dql97lo0aJ6n2NjY9GnTx+L928PeCYnu+ft7Y1Tp06ZDDmZxwdvpKQnT57g+vXrFt9PL168OEsTKzoChpzsXsmSJc32E1+6dClSU1N1nydMmIACBQqgWLFi8PX1xejRo7N9zP79++Pff/+1qL72hpfr5PCcnJxw7do1FC5cGEuXLkX//v0xcuRI1KlTB8eOHcO0adPw7bffGr23NnULEBMTg2LFiul+uTz7oM/R8MEbObxnz1Pz5s3DuHHj8NFHHwEAWrVqBR8fH8ydO9doyOPi4tCsWTPUrl1bb3+TJ09G27Zt7W66K4vYqhUOUXaEhYVJTEyM0XXPtlorXLiwHDt2TG/9pUuXJH/+/EbLbt++XQIDA+XLL7/UW84JD4lyyd69e00u37NnD4KCggAYNmyxtEFLo0aNcOjQIXTt2hW7du3C0qVLM53u2NEw5GRXQkNDTTZs6dmzJwDjDVsyhogSERw9ehTVq1fXrTt//jyKFy9u8pgBAQHYuXMnxowZg2rVqmH58uVZGn3GUTDkZFfKly+PcuXK4YsvvkC+fPkAPA1u2bJlsWnTJpQtW9agjDUatGi1Wnz22WeoX78+OnXq5NDDPT2PT9fJrjx8+BADBw7E+fPnsXLlSt3TbxcXF5w6dUo3h1tu+ueff3D06FE0btzY5OW/I+F7crIrHh4eutdgdevWxc8//2zxvixp0HLz5k24ubmhXbt2SgQcYMjJTvXv3x+bN2/GuHHj0K9fP4vukbPaoOXmzZvo27cvChQogCJFiqBw4cLw8/PDgAEDcOvWLUuqb1d4T052q1q1ajh27Bj69u0LT09PODk5Gd3OVIOWtLQ0NG3a1GyDlpSUFNSvXx/Xrl3DO++8g8qVK0NEcObMGfz444/Yv38/jh49anTmFkfBkJNd8/HxwcqVK81uk5MGLXPnzsXdu3dx+vRpgyfwY8aMwWuvvYZ58+Zh2LBhOfoetsQHb2SX7t+/j23btiE6OhoajQbBwcFo1KgRvLy8DLbdsWMHevbsiWHDhmHEiBG65Vl5WBcaGoqOHTtiyJAhRtfPmjULq1atwu7du3P8nWzGZs1wiExYs2aN+Pn5iUaj0fspWLCgREREGC3zzz//SIMGDaRVq1Zy69YtEclaq7XChQvLqVOnTK4/ffq0FCpUyOLvYg/44I3sysWLF/H2228jNDQUhw8fRnJyMh4+fIhDhw4hNDQUXbp0wcWLFw3KZTRoCQkJQbVq1bB3794sPaxLTExEoUKFTK739/d3+C6nvFwnuzJkyBBcvnwZW7duNbq+RYsWKFOmDGbNmmVyH7///jt69+6Nmzdv4syZM2Yv17VaLW7cuGEy6Ddu3EBAQIBDN47hgzeyK3v37sWUKVNMrh88eDDGjh1rdh+tWrXC8ePHcfTo0SzNYTZ06FC4u7sbXZecnJxpeXvHMznZlfz58+PEiRMmX4tFR0fjlVdeMXsJffPmTWg0GhQsWDDT42W0lc/M801nHQlDTnbl2QEgjLlx4waKFStm0Kvs5s2bGD16NFavXo179+4BePoLo1OnTpg6dWqWAq8qXq6T3cnszPr8eSmnDVru3bsHLy8vaLX6z6HT09ORlJQEHx+fnH0hW7Pdg30iQxqNRrRabaY/z/r6668lICBA4uLiDPZ39epVCQwMlOnTpxs93tq1a6V06dKSlJRksC4pKUnKlCkjv//+u1W+m63wTE52ZdGiRdkus27dOowZM8Zon/HAwECMGjUKq1atMtpqbd68eRg+fLjRziienp4YMWIEZs+ejZYtW2a7XvaC9+Tk8IoUKYJt27ahSpUqRtefOXMGjRo1Qnx8vMG6EiVKYMuWLahQoYLRshcvXkSTJk1w9epVq9Y5L7ExDNmVO3fuYObMmUafnmesy3iwliEnDVri4+Ph7e1tsmy+fPmQkJCQxdrbJ4ac7MrcuXOxfft25M+f32Cdr68vdu7ciXnz5uktT0tL082TZoxWqzU5xluRIkXw119/mSx78eJFFClSJIu1t0+8XCe7UrNmTXzyyScm74E3btyI8ePH4+jRo7plWq0WnTt3NtugZeXKlUZbrfXu3RsJCQlYv3690bJhYWEICgqy6FmBvWDIya4UKFAAp06dQokSJYyuj42NRbVq1ZCYmKhblpMGLX/99ReqV6+OunXrYsKECbomsOfOncPkyZOxb98+HDt2DC+//LKF38j2+HSd7EpaWhoePXpkcn1iYiIeP36stywn3UDLli2L3377Dd26dUO9evX01hUpUgTr1q1z6IADDDnZmXLlymHfvn0IDg42uv7gwYMoV66cwfKcNGgJCwtDVFQUtm7dikuXLkFEUKFCBTRp0sShR4TRseE7eiID06ZNkxIlSkhUVJTBuujoaClSpIhMmzZNb3lOGrRMmDBBTp8+bZ3K2ynek5NdSU1NRb169XD27Fl069ZN7x55xYoVCAkJwb59++Dq6qor06JFC7Rq1crk6C5z587Fhg0b8Pvvvxusa9myJXbu3InixYsjPDwcHTt21E3UoAxb/5Yhel5SUpIMGzZM/P39daPC+Pv7y7Bhw+T+/fsG2xcvXlzOnz9vcn8XLlyQwMBAk+vv3bsnK1askA4dOoinp6cEBgbK0KFDZc+ePZKenm6V72RLPJOTXUtISICImB2M0c3NDZGRkQgMDDS6Pjo6GhUrVkRKSkqmx0tJScHGjRsRERGBDRs2IF++fGjXrh06dOiARo0amX0fb6/YGIbsWqFChTKdPtiaDVrc3NwQHh6OZcuWIT4+HosXLwYAdO/eHZ9++mmW621PeCYnh5cXDVrS09Nx+/Zt+Pv7W7wPW+GZnBzemDFjsHv3bjRr1gx//vkn7ty5gzt37mD//v1o3rw5jhw5gtGjR2d5f/v370dqaqreMq1W65ABB3gmJ0Xs2rUL3bp1w/Xr1/WWFylSBMuWLUOjRo2yvC8fHx+cPHnS5BBUjsbxniIQGWHNBi2qnfcYcnJ4EydORIcOHVC5cmW0bdvW1tWxOww5ObzDhw9j2rRpFjdoWbp0qd7ntLQ0rFmzRu+pfvfu3a1W37zGe3JSwv3797F+/XqsWbMGmzdvRoECBRAeHo4OHTqgXr16ZnuplSpVSu9zXFwcAgICdO/ENRqN0RlRHQVDTsrJaYMWb29vnDp1SpkHb3yFRspRsUFLTvBMTi+MrDZo4ZmcyAHkpEFLyZIl4eLikltVy3M8k5OSVGvQkhM8k5OSeO76Pww5kRENGzZEbGysrathFWwMQ0qwtEHL3r17je5v79692LNnD4KCggAA9evXt15l8xjvyUkJljZo0Wq10Gg0Zi/vNRqN0THbHQVDTkrK6muwihUroly5cvjiiy+QL18+AE/v58uWLYtNmzahbNmyAJ4+cXdUvCenF9rRo0fh4+ODt99+G0+ePEHJkiV1l+gBAQEoWbKkQwccYMjpBefh4YGlS5eif//+qFu3Ln7++WdbV8nq+OCNlJTdBi39+/dH7dq10alTJ+zYsSNL0y45Cp7JSUlnz55F8eLFs1WmWrVqOHbsGO7evQtPT084OTnlUu3yFh+8ESmOl+ukhNjYWBw/fhx169ZFkSJF8Pfff2Pu3LlIT09H+/btUadOHbPl79+/j23btiE6OhoajQbBwcFo1KgRvLy88ugb5KK8msWBKLfs2LFD3NzcRKPRiJ+fnxw/flwCAgIkODhYypUrJy4uLibnQhMRWbNmjfj5+elma8n4KViwoEREROTdF8klDDk5vHr16snAgQPl3r17Mm3aNAkMDJR+/frp1g8dOlRef/11o2UvXLggrq6uEh4eLkeOHJGUlBRJTk6Ww4cPS4cOHcTV1VUuXLiQV18lV/CenByer68vDh8+jHLlyuHRo0fw8PDAoUOHUKNGDQBPZ1CpW7cubt++bVB2yJAhuHz5MrZu3Wp03y1atECZMmUwa9asXP0OuYlP18nhPfu6y9irL3Ovw/bu3WtyNlQAGDx4MPbs2ZOzCtoYQ04Or2rVqpgxYwYePHiAb775BoGBgViwYIFu/fz581GhQgWjZWNjY1GpUiWT+w4JCUFcXJzV65yXeLlODm/37t1o1aoVUlJS4Obmhl27dqFPnz549OgRtFotIiMjsXr1aqNjsjs5OeHatWsmJ1W8ceMGihUrhrS0tNz+GrmGISclxMXF4ejRo6hZsyZKlCiBmzdvYubMmUhOTkbbtm1Rr149o+WcnJxw/fp1FCpUyOj6GzduICAggL3QiBxVRlfTzDhyyNkYhpRhSYOWnExn7Ch4JiclREREoG/fvkhMTNRb7ufnhwULFqB9+/a2qZgd4NN1cngXL17E22+/jdDQUBw+fBjJycl4+PAhDh06hNDQUHTp0gUXL140WvbOnTuYOXMm7t69a3LdvXv3cvsr5Cqeycnh5aRBy6effooDBw7gt99+M1q2ffv2qFu3LkaOHGnVOuclnsnJ4eWkQcvq1asxcOBAk2X79++PlStX5riOtsSQk8PLSYOWqKioTMtGRUXluI62xJCTw0tKSjL7BN3T0xMPHjwwui4tLQ2PHj0yWTYxMRGPHz/OcR1tiSEnJWT2rtvUo6dy5cph3759JssdPHgQ5cqVy1HdbI0hJ4cnInjppZfg5ORk9CcgIMBk2bfeegsTJkwwOib7lStXMHHiRHTu3Dk3q5/r+HSdHN6SJUuytF2PHj0MlqWmpqJevXo4e/YsunXrhooVKwIAzp07hxUrViAkJAT79u2Dq6urVeuclxhyeuE9ePAAH3/8MZYvX45bt24BAAoWLIhu3bph8uTJDj8EFENODu/OnTtYunQpevTogfz58xtd17NnT/j4+GS6r4SEBIiIyV5pjoj35OTw5s6di+3btxsEHHg6aszOnTsxb968LO2rUKFCSgUcYMhJAS9Cg5acYMjJ4b0IDVpygiEnh/ciNGjJCYacHN6L0KAlJxhycngvQoOWnOArNHJ4L0KDlpxgyEkJqjdoyQmGnJSjYoOWnGDIiRTHB29EimPIiRTHkBMpjiEnUhxDTqQ4hpxIcQw5keIYciLFMeREimPIiRTHkBMpjiEnUhxDTqQ4hpxIcQw5keIYciLFMeREimPIiRTHkBMpjiEnUhxDTqQ4hpxIcQw5keIYciLFMeREimPIiRTHkBMpjiEnUtz/Axz87gZa6bFYAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 2700x900 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Visualize the One-hot encoded Prediction Labels\n",
    "plt.figure(figsize=(9,3), dpi=300)\n",
    "plt.imshow(y_train[:50])\n",
    "plt.xticks(np.arange(len(categories)), categories, rotation=90, fontsize=3.5)\n",
    "plt.yticks(fontsize=4)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2) Model Implementation <a id=\"Model\"></a>\n",
    "Let's create a dict gathering every result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 297,
   "metadata": {},
   "outputs": [],
   "source": [
    "Train_dict = {}\n",
    "Test_dict = {}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## A) Decision Trees <a id=\"DecTrees\"></a>\n",
    "\n",
    "Decision Trees are the first model tested as they are usually easy to fine-tune and they can lead high accuracy in classification task. \n",
    "\n",
    "### a) XGBoost <a id=\"XGBoost\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "XGBoost is an optimized implementation of a Gradient Boosting Decision Trees (GBDT). GBDT is usually the most powerful decision trees algorithm as it is specifically designed to reclassify wrongly classify data. However, it may lead to overfitting on the training set. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 300,
   "metadata": {
    "id": "b8mvigLP7Sej"
   },
   "outputs": [],
   "source": [
    "def Grid_search_XGBoost(X_train, X_test, y_train, y_test, class_weight, eval_metric=accuracy_score, printing=True,\n",
    "                        XGB_param={\"n_estimators\": 100, \"max_depth\":5, \"subsample\":0.9, \"colsample_bytree\":0.4,\n",
    "                                   \"gamma\":0,\n",
    "                                   \"reg_alpha\":0,\n",
    "                                   \"reg_lambda\":0,\n",
    "                                   \"eta\":1e-2,\n",
    "                                   \"min_child_weight\":2}):\n",
    "    \n",
    "    \n",
    "    \"\"\"\n",
    "    Return Train accuracy and Test Accuracy for the XGBoost model. \n",
    "    \n",
    "    Parameters\n",
    "    ---------\n",
    "    X_train(ndarray):(n_samples, n_feat) features array, training split\n",
    "    X_test(ndarray):(n_samples, n_feat) features array, test split\n",
    "    y_train(ndarray):(n_samples, n_class) label array, must be one-hot-encoded, training split\n",
    "    y_test(ndarray):(n_samples, n_class) label array, must be one-hot-encoded, test split\n",
    "    class_weight(ndarray):(n_class,) weight for each class \n",
    "    eval_metric(sklearn_function): must be 'accuracy_score' or 'roc_auc_score' \n",
    "    printing(bool): whether or not to print the result \n",
    "    XGB_param(dict): must have the same shape as the one provided as default value. \n",
    "\n",
    "    Output\n",
    "    -------\n",
    "    train_acc_all: training accuracy. \n",
    "    test_acc_all: test accuracy. \n",
    "    \"\"\"\n",
    "\n",
    "    \n",
    "    xgbc = XGBClassifier(n_estimators=XGB_param[\"n_estimators\"], \n",
    "                         max_depth=XGB_param[\"max_depth\"], \n",
    "                         subsample=XGB_param[\"subsample\"], \n",
    "                         colsample_bytree=XGB_param[\"colsample_bytree\"], \n",
    "                         reg_alpha=XGB_param[\"reg_alpha\"],\n",
    "                         reg_lambda=XGB_param[\"reg_lambda\"],\n",
    "                         eval_metric=eval_metric, \n",
    "                         gamma=XGB_param[\"gamma\"],\n",
    "                         eta=XGB_param[\"eta\"],\n",
    "                         min_child_weight=XGB_param[\"min_child_weight\"],\n",
    "                         objective=\"multi:softmax\",\n",
    "                         num_class=10)\n",
    "    \n",
    "    if class_weight is None:\n",
    "        xgbc.fit(X_train, y_train.argmax(axis=1))\n",
    "    else: \n",
    "        sample_weight = (y_train * class_weight).max(axis=1)\n",
    "        xgbc.fit(X_train, y_train.argmax(axis=1), sample_weight=sample_weight)\n",
    "        \n",
    "    if eval_metric == roc_auc_score:\n",
    "        train_acc_all = roc_auc_score(y_train, xgbc.predict_proba(X_train), multi_class=\"ovr\")\n",
    "        test_acc_all = roc_auc_score(y_test, xgbc.predict_proba(X_test), multi_class=\"ovr\")\n",
    "    else: \n",
    "        train_acc_all = accuracy_score(y_train.argmax(axis=1), xgbc.predict(X_train))\n",
    "        test_acc_all = accuracy_score(y_test.argmax(axis=1), xgbc.predict(X_test))\n",
    "    if printing:\n",
    "        print(f\"Training accuracy: {train_acc_all:0.04f}\")\n",
    "        print(f\"Test accuracy: {test_acc_all:0.04f}\")\n",
    "    return train_acc_all, test_acc_all"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For Fine-tuning, please refer to __[XGBClassifier doc](https://xgboost.readthedocs.io/en/stable/parameter.html)__.\n",
    "In particular to reduce overfitting: \n",
    "\n",
    "decrease: \n",
    "+ \"n_estimators\" (the number of trees)\n",
    "+ \"max_depth\" (the maximum depth of each trees)\n",
    "+ \"subsample\" (the fraction of data used for each trees)\n",
    "+ \"colsample_bytree\" (the fraction of feature used for each trees)\n",
    "\n",
    "or increase:\n",
    "+ \"reg_alpha\" (l1 reg)\n",
    "+ \"reg_lambda\" (l2 reg)\n",
    "+ \"eta\" (learning rate)\n",
    "\n",
    "Finally, \n",
    "+ \"min_child_weight\" is an important parameter which quantify the amount of weight needed in a node before to furhter partition. This is an interesting parameter especially when weight are given and can help increase the sensitivity to some class. \n",
    "\n",
    "Here are the best result obtained with playing with these parameter. Considering the amount of parameter, an actual grid search would probably have take a large amount of time and go beyond the scope of the project. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 302,
   "metadata": {
    "id": "Gk1sfdNe76Kl",
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training accuracy: 0.9946\n",
      "Test accuracy: 0.8214\n"
     ]
    }
   ],
   "source": [
    "train_acc_all, test_acc_all = Grid_search_XGBoost(X_train, X_test, y_train, y_test, \n",
    "                                                  class_weight=None, eval_metric=accuracy_score, printing=True,\n",
    "                                                  XGB_param={\"n_estimators\": 100, \"max_depth\":5, \n",
    "                                                             \"subsample\":0.9, \"colsample_bytree\":0.4,\n",
    "                                                             \"gamma\":0,\n",
    "                                                             \"reg_alpha\":0,\n",
    "                                                             \"reg_lambda\":0,\n",
    "                                                             \"eta\":1e-2,\n",
    "                                                             \"min_child_weight\":2})\n",
    "Train_dict.update({\"XGBoost (1)\": train_acc_all})\n",
    "Test_dict.update({\"XGBoost (1)\": test_acc_all})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 304,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training accuracy: 0.9179\n",
      "Test accuracy: 0.8571\n"
     ]
    }
   ],
   "source": [
    "train_acc_all, test_acc_all = Grid_search_XGBoost(X_train, X_test, y_train, y_test, \n",
    "                                                  class_weight=class_weight, eval_metric=accuracy_score, printing=True,\n",
    "                                                  XGB_param={\"n_estimators\": 40, \"max_depth\":3, \n",
    "                                                             \"subsample\":0.9, \"colsample_bytree\":0.1,\n",
    "                                                             \"gamma\":0,\n",
    "                                                             \"reg_alpha\":0,\n",
    "                                                             \"reg_lambda\":0,\n",
    "                                                             \"eta\":1e-1,\n",
    "                                                             \"min_child_weight\":1})\n",
    "\n",
    "Train_dict.update({\"XGBoost (2)\": train_acc_all})\n",
    "Test_dict.update({\"XGBoost (2)\": test_acc_all})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 305,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training accuracy: 0.9357\n",
      "Test accuracy: 0.8214\n"
     ]
    }
   ],
   "source": [
    "train_acc_all, test_acc_all = Grid_search_XGBoost(X_train, X_test, y_train, y_test, \n",
    "                                                  class_weight=None, eval_metric=accuracy_score, printing=True,\n",
    "                                                  XGB_param={\"n_estimators\": 40, \"max_depth\":3, \n",
    "                                                             \"subsample\":0.9, \"colsample_bytree\":0.1,\n",
    "                                                             \"gamma\":0,\n",
    "                                                             \"reg_alpha\":0,\n",
    "                                                             \"reg_lambda\":0,\n",
    "                                                             \"eta\":1e-2,\n",
    "                                                             \"min_child_weight\":2})\n",
    "Train_dict.update({\"XGBoost (3)\": train_acc_all})\n",
    "Test_dict.update({\"XGBoost (3)\": test_acc_all})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### b) Random Forest  <a id=\"RandomForest\"></a>\n",
    "\n",
    "Observing that XGBoost is overfitting, Random Forest is a common decision trees algorithm to address these issues. Adverserly, it may lead to underfitting.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 306,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Grid_search_RandomForest(X_train, X_test, y_train, y_test, printing=True,\n",
    "                             RF_param={\"n_estimators\": 60, \"max_depth\":4, \"min_samples_split\":2, \n",
    "                                       \"min_samples_leaf\":3, \"class_weight\":\"balanced\", \n",
    "                                       \"min_weight_fraction_leaf\":0,\n",
    "                                       \"max_samples\":None}):\n",
    "    \n",
    "    \n",
    "    \"\"\"\n",
    "    Return Train accuracy and Test Accuracy for the RandoForest model. \n",
    "    \n",
    "    Parameters\n",
    "    ---------\n",
    "    X_train(ndarray):(n_samples, n_feat) features array, training split\n",
    "    X_test(ndarray):(n_samples, n_feat) features array, test split\n",
    "    y_train(ndarray):(n_samples, n_class) label array, must be one-hot-encoded, training split\n",
    "    y_test(ndarray):(n_samples, n_class) label array, must be one-hot-encoded, test split\n",
    "    class_weight(ndarray):(n_class,) weight for each class \n",
    "    printing(bool): whether or not to print the result \n",
    "    RF_param(dict): must have the same shape as the one provided as default value. \n",
    "\n",
    "    Output\n",
    "    -------\n",
    "    train_acc_all: training accuracy. \n",
    "    test_acc_all: test accuracy. \n",
    "    \"\"\"\n",
    "    \n",
    "    rf = RandomForestClassifier(n_estimators=RF_param[\"n_estimators\"], \n",
    "                                  criterion=\"log_loss\",\n",
    "                                  max_depth=RF_param[\"max_depth\"], \n",
    "                                  min_samples_split=RF_param[\"min_samples_split\"], \n",
    "                                  min_samples_leaf=RF_param[\"min_samples_leaf\"], \n",
    "                                  class_weight=RF_param[\"class_weight\"],\n",
    "                                  min_weight_fraction_leaf=RF_param[\"min_weight_fraction_leaf\"],\n",
    "                                  max_samples=RF_param[\"max_samples\"])\n",
    "\n",
    "    rf.fit(X_train, y_train.argmax(axis=1))\n",
    "\n",
    "    train_acc_all = accuracy_score(y_train.argmax(axis=1), rf.predict(X_train))\n",
    "    test_acc_all = accuracy_score(y_test.argmax(axis=1), rf.predict(X_test))\n",
    "    \n",
    "    if printing:\n",
    "        print(f\"Training accuracy: {train_acc_all:0.04f}\")\n",
    "        print(f\"Test accuracy: {test_acc_all:0.04f}\")\n",
    "    return train_acc_all, test_acc_all"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For Fine-tuning, please refer to __[RandomForest doc](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.RandomForestClassifier.html)__.\n",
    "In particular to reduce overfitting: \n",
    "\n",
    "decrease: \n",
    "+ \"n_estimators\" (the number of trees)\n",
    "+ \"max_depth\" (the maximum depth of each trees)\n",
    "\n",
    "\n",
    "or increase:\n",
    "+ \"reg_alpha\" (l1 reg)\n",
    "+ \"reg_lambda\" (l2 reg)\n",
    "+ \"eta\" (learning rate)\n",
    "+ \"min_samples_split\" (the amount of data recquired to further partition a tree)\n",
    "+ \"min_samples_leaf\" (the amount of data recquired to further to consider a lead and not supress it)\n",
    "+ \"min_weight_fraction_leaf\" (The same spirit but with weight) \n",
    "\n",
    "Finally, \n",
    "+ \"class_weight\" is an important parameter which allow to deals with the imbalance of the dataset by adding a weight to each class. \n",
    "\n",
    "Here are the best result obtained with playing with these parameter. Considering the amount of parameter, an actual grid search would probably have take a large amount of time and go beyond the scope of the project.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 311,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training accuracy: 0.8643\n",
      "Test accuracy: 0.8357\n"
     ]
    }
   ],
   "source": [
    "#The accuracy is not determinist here. For the Test split, it ranges from 0.78-0.83\n",
    "train_acc_all, test_acc_all = Grid_search_RandomForest(X_train, X_test, y_train, y_test, printing=True,\n",
    "                                                       RF_param={\"n_estimators\": 60, \"max_depth\":4, \n",
    "                                                                 \"min_samples_split\":2, \"min_samples_leaf\":3,  \n",
    "                                                                 \"class_weight\":\"balanced\", \n",
    "                                                                 \"min_weight_fraction_leaf\":0,\n",
    "                                                                 \"max_samples\":None})\n",
    "Train_dict.update({\"RandomForest\": train_acc_all})\n",
    "Test_dict.update({\"RandomForest\": test_acc_all})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## B) Bayes Model <a id=\"Bayes\"></a>\n",
    "\n",
    "Again for simplicity purpose, Bayes model are used as no fine-tuning is recquired and as they may lead to fairly good accuracy depending on the data set. \n",
    "\n",
    "Bayes model are very simple model where class are determined using probability bayes theory. In particular, the probability for a cell to be classified as $y$ knowing its features $x_1,...x_n$ is: \n",
    "$$P(y | x_1,...,x_n) = \\frac{P(x_1,...,x_n | y) * P(y)}{P(x_1,...,x_n)}$$\n",
    "\n",
    "Here $P(y)$ is the prior probability and is simply defined as the proportion of each class y: \n",
    "$$\\frac{1}{count(y)}[\\sum_{i=0}^{num_{class} - 1} 1 * (y==i)]$$\n",
    "\n",
    "$P(x_1,...,x_n | y)$ is the posterior probability knowing $y$ and depends on the Bayes Model selected.\n",
    "\n",
    "The class is finally determined by maximizing the likelihood: \n",
    "\n",
    "$$y_{pred} = argmax_{i= 0:num_{class}}[P(x_1,...,x_n | y == i) * P(y == i)] $$\n",
    "\n",
    "Note, a Naive Bayes Model, each features are assumed to be independent: \n",
    "\n",
    "$$P(x_1,...,x_n | y) = \\prod_{i = 1}^{n} P(x_i | y) $$\n",
    "\n",
    "### a) Gaussian Naive Bayes <a id=\"Gaussian\"></a>\n",
    "\n",
    "For a the Gaussian Naive Bayes Model: \n",
    "$$P(x_i|y) = \\frac{1}{\\sqrt{2 \\pi \\sigma_y}} * exp(-\\frac{(x_i-\\mu_y)^2}{2 \\sigma_y^2})$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 313,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Accuracy: 0.9321428571428572\n",
      "Test Accuracy: 0.7\n"
     ]
    }
   ],
   "source": [
    "nb = GaussianNB()\n",
    "nb = nb.fit(X_train, y_train.argmax(axis=1))\n",
    "train_acc = accuracy_score(y_train.argmax(axis=1), nb.predict(X_train))\n",
    "test_acc = accuracy_score(y_test.argmax(axis=1), nb.predict(X_test))\n",
    "print(f\"Train Accuracy: {train_acc}\")\n",
    "print(f\"Test Accuracy: {test_acc}\")\n",
    "\n",
    "Train_dict.update({\"GNB\": train_acc})\n",
    "Test_dict.update({\"GNB\": test_acc})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### b) Bernouilli Naive Bayes <a id=\"Bernouilli\"></a>\n",
    "\n",
    "For a the Bernouilli Naive Bayes Model: \n",
    "$$P(x_i|y) = P(x_i=1|y)*x_i + P(x_i=0|y)*(1-x_i)$$\n",
    "\n",
    "Here x_i are supposed to be binary. The problem is that the features we are working with are continuous. Unconsciously, the Bernouilli Naives Bayes has been applied and still work. \n",
    "I don't have a perfect reason why the classification could still happen. My assumption is that the posterior probability is actually compute following this formula:  \n",
    "$$P(x_i|y) = P(x_i=x|y)*is(x_i = x) + P(x_i \\neq x|y)*is(x_i \\neq x)$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 316,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Accuracy: 0.9017857142857143\n",
      "Test Accuracy: 0.8357142857142857\n"
     ]
    }
   ],
   "source": [
    "nb = BernoulliNB()\n",
    "\n",
    "nb = nb.fit(X_train, y_train.argmax(axis=1))\n",
    "train_acc = accuracy_score(y_train.argmax(axis=1), nb.predict(X_train))\n",
    "test_acc = accuracy_score(y_test.argmax(axis=1), nb.predict(X_test))\n",
    "print(f\"Train Accuracy: {train_acc}\")\n",
    "print(f\"Test Accuracy: {test_acc}\")\n",
    "\n",
    "Train_dict.update({\"BNB\": train_acc})\n",
    "Test_dict.update({\"BNB\": test_acc})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### c) Multivariate Gaussian Bayes <a id=\"MultiGaussian\"></a>\n",
    "\n",
    "\n",
    "For a the Multivariate Gaussian Bayes Model: \n",
    "$$P(X = x_1,...x_n|y) = \\frac{1}{\\sqrt{(2 \\pi)^k det(\\Sigma_y)}} * exp(-\\frac{1}{2}(X-\\mu_y)^T\\Sigma_y^{-1}(X-\\mu_y))$$\n",
    "\n",
    "Note, $k$ is the rank of the covariance matrix $\\Sigma$.\n",
    "\n",
    "Note that $\\Sigma$ is the Covariance matrix and may not be invertible. Therefore, Scipy allow matrix to not be invertible using the parameter: ```allow_singular = True``` and by using an approximation of both the determinent and the inversion operation. \n",
    "Please refer to the documentation: __[Scipy multivariate_normal](https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.multivariate_normal.html)__."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 324,
   "metadata": {},
   "outputs": [],
   "source": [
    "class MultivariateGaussianBayes():\n",
    "    '''class implementing the Multivariate Gaussian Bayes model'''\n",
    "    def __init__(self):\n",
    "        self.n = 0\n",
    "        self.mu = []\n",
    "        self.cov = []\n",
    "    \n",
    "    def fit(self, X, y):\n",
    "        self.n = len(np.unique(y))\n",
    "        for i in range(self.n):\n",
    "            self.mu.append(X[y==i].mean(axis=0))\n",
    "            self.cov.append(np.cov(X[y==i].T))\n",
    "        self.mu = np.array(self.mu)\n",
    "        self.cov = np.array(self.cov)\n",
    "        self.prior = np.bincount(y) / y.shape[0]\n",
    "    \n",
    "    def predict(self, X):\n",
    "        pdf = []\n",
    "        for i in range(10):\n",
    "            pdf.append(self.prior[i] * multivariate_normal.pdf(X, mgb.mu[i], mgb.cov[i], allow_singular=True))\n",
    "        pdf = np.array(pdf)\n",
    "        return pdf.argmax(axis=0)\n",
    "    \n",
    "    def predict_proba(self, X):\n",
    "        pdf = []\n",
    "        for i in range(10):\n",
    "            pdf.append(self.prior[i] * multivariate_normal.pdf(X, mgb.mu[i], mgb.cov[i], allow_singular=True))\n",
    "        pdf = np.array(pdf)\n",
    "        return pdf.T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 325,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Accuracy: 1.0\n",
      "Test Accuracy: 0.21428571428571427\n"
     ]
    }
   ],
   "source": [
    "mgb = MultivariateGaussianBayes()\n",
    "mgb.fit(X_train, y_train.argmax(axis=1))\n",
    "\n",
    "print(f\"Train Accuracy: {accuracy_score(y_train.argmax(axis=1), mgb.predict(X_train))}\")\n",
    "print(f\"Test Accuracy: {accuracy_score(y_test.argmax(axis=1), mgb.predict(X_test))}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here the model is overfitting a lot. Let's investigate why. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 329,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "exemple of train probability: [[0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
      "  0.00000000e+00 2.44994039e-50 0.00000000e+00 0.00000000e+00\n",
      "  0.00000000e+00 0.00000000e+00]]\n",
      "exemple of train probability: [[0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]]\n"
     ]
    }
   ],
   "source": [
    "print(f\"exemple of train probability: {mgb.predict_proba(X_train)[:1,:]}\")\n",
    "print(f\"exemple of train probability: {mgb.predict_proba(X_test)[:1,:]}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The reason it overfit so much is that probability may be very low for each class other than the true class, leading to python approximating all of them to 0. \n",
    "When looking at the posterior probability, dividing by:\n",
    "$\\frac{1}{\\sqrt{(2 \\pi)^k}}$\n",
    "may explain these small probabilities. Knowing that $k$ is the covariance matrix rank, **probability are directly impacted by the number of features fed**. **A good idea** is therefore to use a **reduction techniques while not removing to much of the variance** in the data as the method compute variance among features. The perfect dimension reduction techniques is therefore the **principal component analysis (PCA)**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 339,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "features kept: 8\n"
     ]
    }
   ],
   "source": [
    "#Let's re-run the processing techniques from the begining but using PCA in addition.\n",
    "categories = df['bulk_labels'].unique()\n",
    "#print(categories)\n",
    "\n",
    "#one-hot encoding\n",
    "y = np.zeros((len(df), len(categories)))\n",
    "for i in range(len(df)):\n",
    "  cell_type = df.iloc[i]['bulk_labels']\n",
    "  pos = np.where(categories == cell_type)[0]\n",
    "  y[i, pos] = 1\n",
    "\n",
    "#remove label when processing input data\n",
    "X = df.drop('bulk_labels', axis=1).values\n",
    "X_pca = PCA(n_components=8).fit_transform(X)\n",
    "print(f\"features kept: {X_pca.shape[1]}\")\n",
    "#shufle and 80:20 split\n",
    "np.random.seed(100)\n",
    "permutation = np.random.permutation(len(X_pca))\n",
    "X_pca, y = X_pca[permutation], y[permutation]\n",
    "\n",
    "X_pca_train, y_train = X_pca[:int(len(X_pca)*0.8)], y[:int(len(y)*0.8)]\n",
    "X_pca_test, y_test = X_pca[int(len(X_pca)*0.8):], y[int(len(y)*0.8):]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 340,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Accuracy: 0.85\n",
      "Test Accuracy: 0.8357142857142857\n"
     ]
    }
   ],
   "source": [
    "mgb = MultivariateGaussianBayes()\n",
    "mgb.fit(X_pca_train, y_train.argmax(axis=1))\n",
    "\n",
    "\n",
    "train_acc = accuracy_score(y_train.argmax(axis=1), mgb.predict(X_pca_train))\n",
    "test_acc = accuracy_score(y_test.argmax(axis=1), mgb.predict(X_pca_test))\n",
    "print(f\"Train Accuracy: {train_acc}\")\n",
    "print(f\"Test Accuracy: {test_acc}\")\n",
    "\n",
    "Train_dict.update({\"MGB\": train_acc})\n",
    "Test_dict.update({\"MGB\": test_acc})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## C) Feed Forward Neural Network <a id=\"FFNN\"></a>\n",
    "\n",
    "Finally a Feed Forward Neural Network (FFNN) is used as neural network offer more freedom in term of architecture and as they can learn non linear pattern in the data. The underlying problem with FFNN is that their fine-tuning may be very tricky as several layers can be considered or regularisation technique etc. Here for simplification purpose, no scheduler for the learning parameter will be implemented."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 342,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "device(type='cuda')"
      ]
     },
     "execution_count": 342,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Enable the use of cuda if run with a GPU.\n",
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "device"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 343,
   "metadata": {},
   "outputs": [],
   "source": [
    "class RNADataset(Dataset):\n",
    "    def __init__(self, X, y):\n",
    "        super().__init__()\n",
    "        self.X = torch.tensor(X, dtype=torch.float).to(device)\n",
    "        self.y = torch.tensor(y.argmax(axis=1), dtype=torch.long).to(device)\n",
    "        \n",
    "    def __len__(self): \n",
    "        return len(self.y)\n",
    "    \n",
    "    def __getitem__(self, idx):\n",
    "        return self.X[idx], self.y[idx]\n",
    "    \n",
    "train_dataset = RNADataset(X_train, y_train)\n",
    "test_dataset = RNADataset(X_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This FFNN architecture is as follow: \n",
    "Base unit repeated $n-2$ times, by providing a list $[n_1,...n_n]$: \n",
    "\n",
    "For $i$ in $[1,...n-2]$\n",
    "+ $Linear(n_i,n_{i+1})$\n",
    "+ $ReLU()$\n",
    "+ $Batchnorm(n_i)$\n",
    "+ $Dropout(p_i)$\n",
    "\n",
    "Then feed to the output layer: \n",
    "+ $Linear(n_{n-1},n_n)$\n",
    "+ $ReLU()$\n",
    "\n",
    "In particular, $(n_1, n_n)$ must respectively be the **input feature number** and **output class number**, which are here **765** and **10**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 345,
   "metadata": {},
   "outputs": [],
   "source": [
    "class FFNN(nn.Module):\n",
    "    def __init__(self, list_emb_dim, list_p):\n",
    "        super().__init__()\n",
    "        self.encod_dim = list_emb_dim\n",
    "        self.list_p = list_p\n",
    "        self.autoencode_layers = [[nn.Linear(self.encod_dim[i], self.encod_dim[i+1]), nn.ReLU(), \n",
    "                                   nn.BatchNorm1d(self.encod_dim[i+1]), \n",
    "                                   nn.Dropout(self.list_p[i])] \n",
    "                                  for i in range(len(self.encod_dim)-2)]\n",
    "        self.autoencoder = nn.Sequential(*[layer for layers in self.autoencode_layers for layer in layers])\n",
    "        self.out = nn.Sequential(nn.Linear(self.encod_dim[-2], self.encod_dim[-1]),\n",
    "                                 nn.ReLU())\n",
    "\n",
    "    def forward(self, X):\n",
    "        X_emb = self.autoencoder(X)\n",
    "        return self.out(X_emb)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 348,
   "metadata": {},
   "outputs": [],
   "source": [
    "class FFNNTrainer:\n",
    "    '''\n",
    "    Model trainer\n",
    "    '''\n",
    "    def __init__(self, model, batch_size, learning_rate, epoch, l1=0, l2=0, weight=None):\n",
    "\n",
    "        '''\n",
    "        Initialization \n",
    "\n",
    "        Parameters\n",
    "        ----------\n",
    "        model(Pytorch model): model to be trained\n",
    "        batch_size(int): size of the batch to use for GraphDataLoader\n",
    "        learning_rate(float): learning rate for the optimiser Adam\n",
    "        epoch(int): The number of training roung to achieve\n",
    "        l1(float): l1 regularization of the loss to reduce overfitting must be positive\n",
    "        l2(float): l2 regularization of the loss to reduce overfitting must be positive\n",
    "        weight(torch.tensor): shape (N_class,) the weight to be used for the cross entropy function.  \n",
    "        '''\n",
    "        self.model = model\n",
    "\n",
    "        num_params = sum(item.numel() for item in model.parameters())\n",
    "        #print(f\"{model.__class__.__name__} - Number of parameters: {num_params}\")\n",
    "\n",
    "        self.batch_size = batch_size\n",
    "        self.optimizer = torch.optim.Adam(self.model.parameters(), learning_rate, weight_decay=l2)\n",
    "        if weight is None:\n",
    "            self.loss = nn.CrossEntropyLoss()\n",
    "        else:\n",
    "            self.loss = nn.CrossEntropyLoss(weight=weight)\n",
    "        self.epoch = epoch\n",
    "        self.l1 = l1\n",
    "        self.classification = nn.Softmax(dim=1)\n",
    "\n",
    "    def train(self, train_dataset, test_dataset, draw_curve=True, early_stop=True):\n",
    "        loader = DataLoader(train_dataset, batch_size=self.batch_size, shuffle=False)\n",
    "\n",
    "        train_loss_all = []\n",
    "        test_loss_all = []\n",
    "        \n",
    "        train_acc_all = []\n",
    "        test_acc_all = []\n",
    "        \n",
    "        train_loss_min = np.inf\n",
    "        test_loss_kept = np.inf\n",
    "        \n",
    "        for i in tqdm(range(self.epoch), leave=True):\n",
    "            epoch_loss = 0\n",
    "            epoch_acc = 0\n",
    "            self.model.train()\n",
    "            for X_batch_train, y_batch_train in loader:\n",
    "                \n",
    "                y_batch_train_pred = self.model(X_batch_train)\n",
    "                \n",
    "                # L1 regularisation\n",
    "                batch_loss = self.loss(y_batch_train_pred, y_batch_train) \n",
    "                batch_loss += self.l1 * torch.norm(torch.cat([x.view(-1) for x in self.model.parameters()]), 1)\n",
    "\n",
    "                self.optimizer.zero_grad()\n",
    "                batch_loss.backward()\n",
    "                self.optimizer.step()\n",
    "\n",
    "                batch_imp = X_batch_train.shape[0] / train_dataset.X.shape[0]\n",
    "                epoch_loss += batch_loss.detach().cpu().numpy() * batch_imp\n",
    "                \n",
    "                y_batch_train_pred_class = self.classification(y_batch_train_pred).argmax(dim=1).detach().cpu().numpy()\n",
    "                epoch_acc += accuracy_score(y_batch_train.detach().cpu().numpy(), y_batch_train_pred_class) * batch_imp\n",
    "                \n",
    "            train_loss_all.append(epoch_loss)\n",
    "            train_acc_all.append(epoch_acc)\n",
    "            \n",
    "            test_loss, test_acc = self.evaluate(test_dataset)\n",
    "            test_loss_all.append(test_loss)\n",
    "            test_acc_all.append(test_acc)\n",
    "            \n",
    "            if early_stop:\n",
    "                if train_loss_all[-1] < train_loss_min:\n",
    "                    train_loss_min = train_loss_all[-1]\n",
    "                    test_loss_kept = test_loss_all[-1]\n",
    "                    weights = self.model.state_dict()\n",
    "                    \n",
    "        if early_stop:\n",
    "            self.model.load_state_dict(weights)\n",
    "            index = np.argmin(train_loss_all)\n",
    "        else:\n",
    "            index = self.epoch - 1\n",
    "        if draw_curve:\n",
    "\n",
    "            fig, ax = plt.subplots(1, 2, figsize=(14, 8), constrained_layout=True)\n",
    "            ax[0].set_title(\"Training vs Test Loss per Epochs\", fontsize=20)\n",
    "            ax[0].plot(np.arange(len(train_loss_all)), train_loss_all, label='train', color='b')\n",
    "            ax[0].plot(np.arange(len(test_loss_all)), test_loss_all, label='test', color='r')\n",
    "            ax[0].set_xlabel(\"# Epochs\", fontsize=20)\n",
    "            ax[0].set_ylabel(\"Loss\", fontsize=20)\n",
    "            ax[0].legend()\n",
    "            ax[0].scatter(index, train_loss_all[index], c='b')\n",
    "            ax[0].scatter(index, test_loss_all[index], c='r')\n",
    "            ax[0].annotate(f\"{train_loss_all[index]:0.03f}\", (index, train_loss_all[index]), fontsize=15)\n",
    "            ax[0].annotate(f\"{test_loss_all[index]:0.03f}\", (index, test_loss_all[index]), fontsize=15)\n",
    "            \n",
    "            ax[1].set_title(\"Training vs Test Accuracy per Epochs\", fontsize=20)\n",
    "            ax[1].plot(np.arange(len(train_acc_all)), train_acc_all, label='train', color='b')\n",
    "            ax[1].plot(np.arange(len(test_acc_all)), test_acc_all, label='test', color='r')\n",
    "            ax[1].set_xlabel(\"# Epochs\", fontsize=20)\n",
    "            ax[1].set_ylabel(\"Accuracy\", fontsize=20)\n",
    "            ax[1].legend()\n",
    "            ax[1].scatter(index, train_acc_all[index], c='b')\n",
    "            ax[1].scatter(index, test_acc_all[index], c='r')\n",
    "            ax[1].annotate(f\"{train_acc_all[index]:0.03f}\", (index, train_acc_all[index]), fontsize=15)\n",
    "            ax[1].annotate(f\"{test_acc_all[index]:0.03f}\", (index, test_acc_all[index]), fontsize=15)\n",
    "            \n",
    "        return train_loss_min, test_loss_kept, train_acc_all[index], test_acc_all[index], self.model\n",
    "\n",
    "    def evaluate(self, test_dataset):\n",
    "        self.model.eval()\n",
    "        with torch.no_grad():\n",
    "            y_batch_test_pred = self.model(test_dataset.X)\n",
    "            test_loss = self.loss(y_batch_test_pred, test_dataset.y) \n",
    "            test_loss += self.l1 * torch.norm(torch.cat([x.view(-1) for x in model.parameters()]), 1)\n",
    "            \n",
    "            y_batch_test_pred_class = self.classification(y_batch_test_pred).argmax(dim=1).detach().cpu().numpy()\n",
    "            test_acc = accuracy_score(test_dataset.y.detach().cpu().numpy(), y_batch_test_pred_class)\n",
    "            \n",
    "        return test_loss.detach().cpu().numpy(), test_acc"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here are the best result obtained with playing with these parameter. Considering the amount of parameter, an actual grid search would probably have take a large amount of time and go beyond the scope of the project.  \n",
    "\n",
    "Let's try a complex architecture with several layer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 351,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 30/30 [00:01<00:00, 17.23it/s]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = FFNN([765, 512, 256, 128, 64, 32, 10], list_p=[0.4, 0.4, 0.2, 0.4, 0.2])\n",
    "model = model.to(device)\n",
    "Trainer = FFNNTrainer(model, batch_size=100, learning_rate=1e-3, epoch=30, l1=0, l2=0)\n",
    "train_loss_all, test_loss_all, train_acc_all, test_acc_all, model = Trainer.train(train_dataset, \n",
    "                                                                                       test_dataset, \n",
    "                                                                                       draw_curve=True, \n",
    "                                                                                       early_stop=True)\n",
    "\n",
    "Train_dict.update({\"FFNN_complex\": train_acc_all})\n",
    "Test_dict.update({\"FFNN_complex\": test_acc_all})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Considering the overfitting, let's try a simpler architecture with less layer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 355,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 20/20 [00:00<00:00, 32.33it/s]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = FFNN([765, 256, 10], list_p=[0.4])\n",
    "model = model.to(device)\n",
    "Trainer = FFNNTrainer(model, batch_size=100, learning_rate=1e-3, epoch=20, l1=1e-3, l2=1e-3)\n",
    "train_loss_all, test_loss_all, train_acc_all, test_acc_all, model = Trainer.train(train_dataset, \n",
    "                                                                                       test_dataset, \n",
    "                                                                                       draw_curve=True, \n",
    "                                                                                       early_stop=True)\n",
    "Train_dict.update({\"FFNN_simple\": train_acc_all})\n",
    "Test_dict.update({\"FFNN_simple\": test_acc_all})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The overfitting is still present, but a simpler architecture lead to similar result faster!\n",
    "Let's try to take into account the imbalancy of the data set by adding class weight."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 356,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([ 0.6202,  0.3333,  2.5806,  1.1765,  0.8421,  1.4815,  4.2105,  1.8605,\n",
       "        10.0000,  6.1538], device='cuda:0')"
      ]
     },
     "execution_count": 356,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weight = torch.tensor(class_weight, dtype=torch.float).to(device) * 10\n",
    "weight"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Again let's compare a complex architecture with a simple architecture. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 357,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 50/50 [00:04<00:00, 10.71it/s]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = FFNN([765, 350, 128, 50, 10], list_p=[0.2, 0.4, 0.4])\n",
    "model = model.to(device)\n",
    "Trainer = FFNNTrainer(model, batch_size=50, learning_rate=1e-3, epoch=50, l1=0, l2=1e-4, weight=weight)\n",
    "train_loss_all, test_loss_all, train_acc_all, test_acc_all, model = Trainer.train(train_dataset, \n",
    "                                                                                       test_dataset, \n",
    "                                                                                       draw_curve=True, \n",
    "                                                                                       early_stop=True)\n",
    "Train_dict.update({\"FFNN_complex_weight\": train_acc_all})\n",
    "Test_dict.update({\"FFNN_complex_weight\": test_acc_all})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 369,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 20/20 [00:00<00:00, 32.90it/s]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = FFNN([765, 256, 10], list_p=[0.1])\n",
    "model = model.to(device)\n",
    "Trainer = FFNNTrainer(model, batch_size=100, learning_rate=1e-3, epoch=20, l1=1e-3, l2=1e-3, weight=weight)\n",
    "train_loss_all, test_loss_all, train_acc_all, test_acc_all, model = Trainer.train(train_dataset, \n",
    "                                                                                  test_dataset, \n",
    "                                                                                  draw_curve=True, \n",
    "                                                                                  early_stop=True)\n",
    "Train_dict.update({\"FFNN_simple_weight\": train_acc_all})\n",
    "Test_dict.update({\"FFNN_simple_weight\": test_acc_all})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3) Conclusion <a id=\"Conclusion\"></a>\n",
    "Let's plot the different accuracy result obtained"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 420,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x900 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_theme()\n",
    "fig = plt.figure(figsize=(15,9))\n",
    "ax = plt.gca()\n",
    "sns.lineplot(Train_dict, ax=ax, c='b', label=\"Train\")\n",
    "sns.lineplot(Test_dict, ax=ax, c='r', label=\"Test\")\n",
    "ax.scatter(np.arange(len(Train_dict)), list(Train_dict.values()), c='b')\n",
    "for i, v in enumerate(list(Train_dict.values())):\n",
    "    ax.annotate(f\"{v:0.03f}\", (i, v), fontsize=12)\n",
    "\n",
    "ax.scatter(np.arange(len(Test_dict)), list(Test_dict.values()), c='r')\n",
    "for i, v in enumerate(list(Test_dict.values())):\n",
    "    ax.annotate(f\"{v:0.03f}\", (i, v), fontsize=12)    \n",
    "    \n",
    "ax.set_xticks(np.arange(len(Train_dict)), list(Train_dict.keys()), rotation=45, fontsize='medium')\n",
    "plt.yticks(fontsize='medium')\n",
    "ax.set_xlabel(\"Model used\", fontsize='x-large')\n",
    "ax.set_ylabel(\"Accuracy score\", fontsize='x-large')\n",
    "\n",
    "ax.set_title(\"Accuracy comparison accross model\", fontsize='xx-large')\n",
    "ax.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "When looking at this graph several conclusion can be made: \n",
    "+ It happened that the **best test accuracy** obtained is with a **Feed Forward Neural Network (FFNN)** and in particular with the variant with a complex architecture and weight. However FFNN are not deterministic, and their accuracy may vary accross test from ~0.83-0.90. FFNN is however the most difficult to fine-tune and lead to a lot of over-fitting. \n",
    "+ **XGBoost** has the advantage to provide a more deterministic solution (accross every test conducted, always the same accuracy for a chosen architecture). In particular the architecture used in XGBoost (2) provide a good trade off in term of Test accuracy and over-fitting. \n",
    "+ The **minimum over-fitting** is observed for the **RandomForest** Classifier and the **Multivariate Gaussian Model (MGB) fed with PCA**. MGB present the advantage to be completely deterministic, when random forest may reach various different accuracy from ~0.80-0.84. However, MGB had to be fed with PCA. Therefore, for larger data set with a bigger number of features (there is usually 24,000 genes instead of only 750 here), this method may not be as powerful.  \n",
    "+ For **Naive Bayes method**, the gaussian one is not good enough, as it overfit too much. Bernouilli is unexpectedly good. It should however be avoided for larger dataset as X features may not be as discrete as in our case.\n",
    "\n",
    "To conclude on the accurcay comparison for each model, FFNN is the most powerful, Decision Trees gives a fairly good result with a minimum amount of fine tuning, Multivariate Gaussian Bayes fed with PCA is the simplest and lead to still good result.\n",
    "To fully assess the accuracy of each method, different generation of train / test split should be tested, and each model should be tested several times. \n",
    "\n",
    "Thank you for reading! "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "Final Project Part 2 - Classification",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Chem142",
   "language": "python",
   "name": "chem142"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}