Running_skewed_model_scikit.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "myZjbbTP_sKi"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib\n",
        "import matplotlib.pyplot as plt\n",
        "from matplotlib import cm\n",
        "from imblearn.datasets import make_imbalance"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import warnings\n",
        "from sklearn.exceptions import DataConversionWarning\n",
        "warnings.filterwarnings('ignore')"
      ],
      "metadata": {
        "id": "QzJ3R6Cb_wBJ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "list_files = np.load('octmnist.npz')\n",
        "for name in list_files:\n",
        "  print(name)\n",
        "  print(np.shape(list_files[name]))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Jini8FhV_xA0",
        "outputId": "4579a1c6-0e42-4590-8bf3-a788bb2aad84"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "train_images\n",
            "(97477, 28, 28)\n",
            "val_images\n",
            "(10832, 28, 28)\n",
            "test_images\n",
            "(1000, 28, 28)\n",
            "train_labels\n",
            "(97477, 1)\n",
            "val_labels\n",
            "(10832, 1)\n",
            "test_labels\n",
            "(1000, 1)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_images = list_files['train_images']\n",
        "train_labels = list_files['train_labels']\n",
        "\n",
        "test_images = list_files['test_images']\n",
        "test_labels = list_files['test_labels']\n",
        "\n",
        "val_images = list_files['val_images']\n",
        "val_labels = list_files['val_labels']"
      ],
      "metadata": {
        "id": "Xu7FmMsj_0At"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "x_train1 = train_images.reshape(97477, -1)\n",
        "y_train1 = train_labels.reshape(97477, -1)\n",
        "\n",
        "x_test1 = test_images.reshape(1000, -1)\n",
        "y_test1 = test_labels.reshape(1000, -1)\n",
        "\n",
        "x_val1 = val_images.reshape(10832, -1)\n",
        "y_val1 = val_labels.reshape(10832, -1)"
      ],
      "metadata": {
        "id": "wzzDZAAk_2zi"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "x_train1 = (255 - x_train1).astype(np.int64)\n",
        "x_test1 = (255 - x_test1).astype(np.int64)\n",
        "x_val1 = (255 - x_val1).astype(np.int64)\n"
      ],
      "metadata": {
        "id": "BkFOLleB_32B"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "x_train1 = x_train1 / 255.\n",
        "x_test1 = x_test1 / 255.\n",
        "x_val1 = x_val1 / 255."
      ],
      "metadata": {
        "id": "_SsY2nzy_5Q8"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.ensemble import AdaBoostClassifier, RandomForestClassifier\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "from sklearn.svm import LinearSVC\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.naive_bayes import MultinomialNB\n",
        "from sklearn.feature_selection import SelectFromModel\n",
        "from sklearn.model_selection import cross_val_predict\n",
        "from sklearn.metrics import accuracy_score, precision_score\n",
        "from sklearn.pipeline import Pipeline\n",
        "from sklearn.decomposition import PCA\n",
        "from sklearn.metrics import confusion_matrix\n",
        "from sklearn.metrics import precision_score, recall_score, f1_score, precision_recall_curve"
      ],
      "metadata": {
        "id": "PNeNmCCfAKfy"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.preprocessing import label_binarize\n",
        "from sklearn.metrics import roc_curve, roc_auc_score, auc"
      ],
      "metadata": {
        "id": "JnN4EdbIAUuw"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "lr_model_final = LogisticRegression(\n",
        "    penalty = 'l2',\n",
        "    C = 0.01,\n",
        "    solver = 'newton-cg',\n",
        "    warm_start = True,\n",
        "    n_jobs = -1,\n",
        "    random_state = 30,\n",
        "    multi_class = 'ovr'\n",
        ")\n",
        "\n",
        "select_lr_final = SelectFromModel(\n",
        "    lr_model_final,\n",
        "    threshold = 'median'\n",
        ")\n",
        "\n"
      ],
      "metadata": {
        "id": "rzmFmhhbAHt2"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "sel_x_train_final_weight = select_lr_final.fit_transform(x_train1, y_train1)\n",
        "sel_x_test_final_weight = select_lr_final.transform(x_test1)\n"
      ],
      "metadata": {
        "id": "fnYeTajq_6oT"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "lr_rf_model_final_weight = RandomForestClassifier(\n",
        "          criterion = 'entropy',\n",
        "          max_depth = None,\n",
        "          n_estimators = 1000,\n",
        "          n_jobs = -1,\n",
        "          random_state = 35,\n",
        "          class_weight = 'balanced'\n",
        "      )"
      ],
      "metadata": {
        "id": "7l7l-O06BspZ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "lr_rf_model_final_weight.fit(sel_x_train_final_weight, y_train1)\n",
        "rf_model1_pred_weight = lr_rf_model_final_weight.predict(sel_x_test_final_weight)"
      ],
      "metadata": {
        "id": "GDLZ0e9J_8x1"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "CM_lr_rf_weight = confusion_matrix(y_test1, rf_model1_pred_weight)\n",
        "\n",
        "CM_lr_rf_weight = CM_lr_rf_weight.astype('float') / CM_lr_rf_weight.sum(axis=1)[:, np.newaxis]\n",
        "\n",
        "np.fill_diagonal(CM_lr_rf_weight,0)\n",
        "plt.figure(1, figsize = (18,18))\n",
        "plt.matshow(CM_lr_rf_weight, cmap=plt.cm.coolwarm)\n",
        "plt.title('Confusion Matrix for LR + RF: Weighted Data')\n",
        "plt.tight_layout\n",
        "plt.colorbar()\n",
        "plt.show()\n",
        "                          "
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 444
        },
        "id": "icagYQij_9na",
        "outputId": "ce3e963a-f557-40d2-eb0d-d76f48fd31bb"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1800x1800 with 0 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 480x480 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAagAAAGZCAYAAADclKXIAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAA/UklEQVR4nO3dfVxUVf4H8M8MMjMiD4oE+ICiYCqp0IIQoGmFsWYmuRXmFkg+ZdCmVJtsKT4llS7hz1S0RF1N5aVl7pqhhZGLYhZGaZZJPkAmD1aCoILOnN8f7kwOAzjDjM69zef9et1XzZl77/neKzPfOeeee65CCCFAREQkMUp7B0BERNQcJigiIpIkJigiIpIkJigiIpIkJigiIpIkJigiIpIkJigiIpIkJigiIpIkJigiIpIk2SWo48eP4/7774eHhwcUCgU++OADm+7/1KlTUCgUWLt2rU33K2fDhw/H8OHDbba/uro6TJo0Cb6+vlAoFJg+fbrN9k3yoVAoMGfOnDZvm5KSYtuALGTrzwWZalOC+vHHHzF16lT07t0bGo0G7u7uiI6OxpIlS3Dp0iVbx2gkMTERhw8fxquvvor169cjLCzsptZ3K02YMAEKhQLu7u7Nnsfjx49DoVBAoVBg8eLFFu//559/xpw5c1BSUmKDaNtu4cKFWLt2LaZNm4b169fjySefvKn1+fv748EHH2x1Hf251y9qtRq33347Zs+ejcuXL9/U+GwRz/XrXr/4+vpaVO8DDzyATp06oekMaF999RUUCgV69uxpss2ePXugUCiwatUqyw7yFti/fz/mzJmD8+fP2y0Gf39/w7+HUqlEx44dMXDgQEyZMgWff/65VfteuHChzX+kS0k7Szf48MMP8eijj0KtViMhIQEDBgxAY2MjCgsL8eKLL+Lbb7+9aX+oly5dQlFREV5++eWb9uupZ8+euHTpEpydnW/K/m+kXbt2uHjxIv7zn//gscceM3rv3XffhUajafMX5s8//4y5c+fC398fISEhZm+3e/fuNtXXkj179uCuu+5Cenq6TfdrLbVajXfeeQcAUFNTg+3bt2P+/Pn48ccf8e6770o+nhEjRiAhIcGorH379hbVOWTIEHz00Uc4cuQIBg4caCjft28f2rVrh7KyMvz000/o3r270Xv6bS1x6dIltGtn8VeQRfbv34+5c+diwoQJ6Nix402tqzUhISF4/vnnAQAXLlzAd999hy1btuDtt9/GjBkzkJmZ2ab9Lly4EI888gji4uJsGK10WPTXcfLkSYwbNw49e/bEnj170KVLF8N7ycnJKC0txYcffmjzIPWqq6sB4Kb+oSkUCmg0mpu2/xtRq9WIjo7Gpk2bTBLUxo0bMWrUKLz33nu3JJaLFy/CxcUFKpXKpvutqqpCUFCQzfZ39epV6HQ6q+Ns164dnnjiCcPrZ555BlFRUdi0aRMyMzPh4+Nj0f5OnTqFXr164dNPP21TV5Cl8dx+++1G67eFPskUFhaaJKgHHngAe/bsQWFhIcaNG2d4r7CwEJ07d0b//v0tqsuen7NbrVu3bib/Nq+//jrGjx+PN998E3369MG0adPsFJ10WdTF98Ybb6Curg6rV682Sk56gYGBeO655wyvr169ivnz5yMgIABqtRr+/v74xz/+gYaGBqPt9F0whYWFCA8Ph0ajQe/evfGvf/3LsM6cOXMM3QsvvvgiFAoF/P39AVzrDtH///XmzJkDhUJhVPbxxx9jyJAh6NixI1xdXdG3b1/84x//MLzf0jWoPXv2YOjQoejQoQM6duyIMWPG4Lvvvmu2vtLSUsMvNg8PDyQlJeHixYstn9gmxo8fj48++sioW+KLL77A8ePHMX78eJP1f/31V7zwwgsYOHAgXF1d4e7ujpEjR+Lrr782rFNQUIDBgwcDAJKSkgxdDvrjHD58OAYMGIDi4mLcfffdcHFxMZyXpn3tiYmJ0Gg0JscfGxuLTp064eeff272uAoKCqBQKHDy5El8+OGHhhhOnToF4FrimjhxInx8fKDRaBAcHIx169YZ7UP/77N48WJkZWUZ/raOHj1q1rm1hEKhwJAhQyCEwIkTJ2y+/1sZT01NDb7//nvU1NS0ul54eDhUKpWhVaS3b98+3H333QgPDzd6T6fT4cCBA4iKijJ81s6fP4/p06fDz88ParUagYGBeP3116HT6UyOp+k1qIKCAoSFhUGj0SAgIAArV65s9nOs98EHH2DAgAFQq9W44447kJeXZ3hvzpw5ePHFFwEAvXr1Mvl7A4ANGzYgNDQU7du3h6enJ8aNG4fy8nKTelatWoWAgAC0b98e4eHh+O9//9vqeTRH+/btsX79enh6euLVV1816lZdvHgxoqKi0LlzZ7Rv3x6hoaHYunWr0fYKhQL19fVYt26d4dgmTJgAADh9+jSeeeYZ9O3bF+3bt0fnzp3x6KOPGh27pZYtWwZ/f39oNBpERETg4MGDra6flZVlqN/Pzw8zZsywuPfHohbUf/7zH/Tu3RtRUVFmrT9p0iSsW7cOjzzyCJ5//nl8/vnnyMjIwHfffYdt27YZrVtaWopHHnkEEydORGJiInJycjBhwgSEhobijjvuwNixY9GxY0fMmDEDjz/+OB544AG4urpaEj6+/fZbPPjggxg0aBDmzZsHtVqN0tJSkw9jU5988glGjhyJ3r17Y86cObh06RKWLl2K6OhoHDp0yCQ5PvbYY+jVqxcyMjJw6NAhvPPOO/D29sbrr79uVpxjx47F008/jffffx9PPfUUgGutp379+uFPf/qTyfonTpzABx98gEcffRS9evVCZWUlVq5ciWHDhuHo0aPo2rUr+vfvj3nz5mH27NmYMmUKhg4dCgBG/5a//PILRo4ciXHjxuGJJ55oscWwZMkS7NmzB4mJiSgqKoKTkxNWrlyJ3bt3Y/369ejatWuz2/Xv3x/r16/HjBkz0L17d0OXx2233YZLly5h+PDhKC0tRUpKCnr16oUtW7ZgwoQJOH/+vNEPHwBYs2YNLl++jClTpkCtVsPT09Osc2sp/Qe6U6dON2X/lmotnsuXL+PcuXNGZW5ublCr1di2bRuSkpKwZs0aw5dYczQaDUJDQ1FYWGgoKy8vR3l5OaKionD+/HmjXpLDhw+jtrbW0PK6ePEihg0bhjNnzmDq1Kno0aMH9u/fj7S0NJw9exZZWVkt1v3VV1/hz3/+M7p06YK5c+dCq9Vi3rx5uO2225pdv7CwEO+//z6eeeYZuLm54f/+7//wl7/8BWVlZejcuTPGjh2LH374AZs2bcKbb74JLy8vADDs79VXX8WsWbPw2GOPYdKkSaiursbSpUtx991346uvvjL01KxevRpTp05FVFQUpk+fjhMnTuChhx6Cp6cn/Pz8Wjwec7i6uuLhhx/G6tWrcfToUdxxxx0Arn3GHnroIfz1r39FY2MjNm/ejEcffRQ7duzAqFGjAADr16/HpEmTEB4ejilTpgAAAgICAFz7Qbt//36MGzcO3bt3x6lTp7BixQoMHz4cR48ehYuLi0Vx5ubmIjU1FdnZ2YiIiEBWVhZiY2Nx7NgxeHt7m6y/ceNGzJw5Ezk5OYiKisIPP/xguK5qUXemMFNNTY0AIMaMGWPW+iUlJQKAmDRpklH5Cy+8IACIPXv2GMp69uwpAIi9e/cayqqqqoRarRbPP/+8oezkyZMCgFi0aJHRPhMTE0XPnj1NYkhPTxfXH+Kbb74pAIjq6uoW49bXsWbNGkNZSEiI8Pb2Fr/88ouh7OuvvxZKpVIkJCSY1PfUU08Z7fPhhx8WnTt3brHO64+jQ4cOQgghHnnkEXHfffcJIYTQarXC19dXzJ07t9lzcPnyZaHVak2OQ61Wi3nz5hnKvvjiC5Nj0xs2bJgAILKzs5t9b9iwYUZlu3btEgDEggULxIkTJ4Srq6uIi4u74TEKce3fe9SoUUZlWVlZAoDYsGGDoayxsVFERkYKV1dXUVtbazguAMLd3V1UVVW1ub6m9Oe+urpaVFdXi9LSUrF48WKhUCjEgAEDhE6nM6uu6+lj/fTTTy3e1tJ4ADS76P+t16xZ0+K/fVMvvviiACB++uknIYQQmzZtEhqNRjQ0NIidO3cKJycnw7/HW2+9JQCIffv2CSGEmD9/vujQoYP44YcfjPY5c+ZM4eTkJMrKyoxiTk9PN7wePXq0cHFxEWfOnDGUHT9+XLRr1040/aoCIFQqlSgtLTWUff311wKAWLp0qaFs0aJFAoA4efKk0fanTp0STk5O4tVXXzUqP3z4sGjXrp2hvLGxUXh7e4uQkBDR0NBgWG/VqlUCgMnnojk3+vvTfy9t377dUHbx4kWjdRobG8WAAQPEvffea1TeoUMHkZiYaLLPptsLIURRUZEAIP71r3/dMOamwsPDRXJysuG1VqsVXbt2FRkZGc2un5ycbBJramqqiI6Otqhes7v4amtrAVz7RWaOnTt3AgBSU1ONyvW/mpteqwoKCjL8qgeu/crp27evTbtW9L+Itm/fbtLd0JKzZ8+ipKQEEyZMMPqVPmjQIIwYMcJwnNd7+umnjV4PHToUv/zyi+EcmmP8+PEoKChARUUF9uzZg4qKima794Br162Uymv/lFqtFr/88ouh+/LQoUNm16lWq5GUlGTWuvfffz+mTp2KefPmYezYsdBoNFi5cqXZdTW1c+dO+Pr64vHHHzeUOTs7429/+xvq6urw2WefGa3/l7/8pcVf1m1VX1+P2267DbfddhsCAwPxwgsvIDo6Gtu3b2+xi+l6dXV1OHfunGH57bffAFzrXru+/EbdbG2NZ8yYMfj444+NltjYWADXusGFEK22nvT0rSF9N9a+ffsQGhoKlUqFyMhIQ7ee/j2NRmMYTbtlyxYMHToUnTp1MjrmmJgYaLVa7N27t9k6tVotPvnkE8TFxRm1wAMDAzFy5Mhmt4mJiTG0GIBrn0l3d3ezvjPef/996HQ6PPbYY0Zx+vr6ok+fPvj0008BAF9++SWqqqrw9NNPG13jnDBhAjw8PG5Yjzn0PUEXLlwwlF0/uOW3335DTU0Nhg4davbn+frtr1y5gl9++QWBgYHo2LGjRd8JANDY2Iji4mLExMQYypRKJWJiYlBUVNTsNlFRUSguLjZ0A544cQI7d+7EAw88YFHdZnfxubu7AzA+ia05ffo0lEolAgMDjcp9fX3RsWNHnD592qi8R48eJvvo1KmT4UNuC/Hx8XjnnXcwadIkzJw5E/fddx/Gjh2LRx55xPAF39xxAEDfvn1N3uvfvz927dqF+vp6dOjQwVDe9Fj03TG//fab4TzeyAMPPAA3Nzfk5uaipKQEgwcPRmBgYLN9yDqdDkuWLMHy5ctx8uRJaLVaw3udO3c2qz7g2oVcSwYaLF68GNu3b0dJSQk2btzYbFPfXKdPn0afPn1M/h30F96b/r306tWrzXW1RKPR4D//+Q8A4KeffsIbb7yBqqoqs0fCpaSkmFwzA2AywmrYsGEoKCiweTzdu3c3+hJpq+joaCgUCuzbtw/jxo3Dvn37MGLECADXfuQFBQUZyvbt24fBgwcb/m6OHz+Ob775psUfD1VVVS2WX7p0yeT7AkCzZYB13xnHjx+HEAJ9+vRp9n39KF79313T9ZydndG7d+8b1mOOuro6AMY//nfs2IEFCxagpKTE5Jr99T90hRBobGw0+fF76dIlZGZmYsOGDSbXhKurq1FbWwshhMkPHbVaDbVabVR27tw5aLVaky5/Hx8ffP/9980e0/jx43Hu3DnDNdOrV6/i6aefNrrebw6LElTXrl1x5MgRiyow55cnADg5OTVbLsx4In1LdVz/RQ1c+1Wxd+9efPrpp/jwww+Rl5eH3Nxc3Hvvvdi9e3eLMVjKmmPRU6vVGDt2LNatW4cTJ060ekPjwoULMWvWLDz11FOYP38+PD09oVQqMX36dLNbioDlQ5K/+uorwxfO4cOHjVo/N5ulsZrDycnJ6As+NjYW/fr1w9SpU/Hvf//7htv//e9/NxqpVVlZiSeeeAKLFy9GcHCwodzc61nWxtNWnTt3Rr9+/VBYWIi6ujp88803RrcEREVFobCwED/99BPKysrw17/+1fCeTqfDiBEj8Pe//73Zfd9+++02i9Oaz5lOp4NCocBHH33U7H4svb5tDf13qj4R//e//8VDDz2Eu+++G8uXL0eXLl3g7OyMd955B7tzt5q03DZt2oRNmzaZXd+7776Ld999F66urobkqJeent7mm6evV1BQgIULF2L58uWIiIhAaWkpnnvuOcyfPx+zZs0yez8WDZJ48MEHsWrVKhQVFSEyMrLVdXv27AmdTofjx48bDT+trKzE+fPnm73hr606derU7I14TX91A9eapvfddx/uu+8+ZGZmYuHChXj55Zfx6aefNvvrUx/nsWPHTN77/vvv4eXlZdR6sqXx48cjJycHSqXSaFhvU1u3bsU999yD1atXG5WfP3/ecGEYMP/Hgjnq6+uRlJSEoKAgREVF4Y033sDDDz9sGCloqZ49e+Kbb76BTqczakXpf6HZ8u/FXF26dMGMGTMwd+5cHDhwAHfddVer6wcFBRkNn9e3dkNDQ20y44Cl8VhjyJAhyMnJwe7du6HVao0G0+iHuutbgdff/xQQEIC6ujqLW3Le3t7QaDQoLS01ea+5MnO19DcfEBAAIQR69erVatLU/90dP34c9957r6H8ypUrOHnypNEPj7aoq6vDtm3b4OfnZ/iefO+996DRaLBr1y6j1syqVavwG7RYp+kNFysnAboIHRLrTqC8vNyoV6dp6wkAvLy84OTkhMrKSqPyysrKFm8EnzVrFp588klMmjQJADBw4EDU19djypQpePnll1vssWrKoqP8+9//jg4dOmDSpEkmwQLXZphYsmQJABj6GpuO2tGP4NCPRLGFgIAA1NTU4JtvvjGUnT171mSk4K+//mqyrf6G1abNaL0uXbogJCQE69atM0qCR44cwe7duy3uU7XEPffcg/nz5+Ott95qdUYAJycnk1+NW7ZswZkzZ4zK9InUFnfVv/TSSygrK8O6deuQmZkJf39/JCYmtngeb+SBBx5ARUUFcnNzDWVXr17F0qVL4erqimHDhlkdc1s8++yzcHFxwWuvvWaX+ptqazzmDjPXGzJkCLRaLRYvXow+ffoYddlFRUWhrq4Oy5cvh1KpNEpejz32GIqKirBr1y6TfZ4/fx5Xr15ttj59a/GDDz4w6pIqLS3FRx99ZO5hmmjpb37s2LFwcnLC3LlzTT47Qgj88ssvAICwsDDcdtttyM7ORmNjo2GdtWvXWv05unTpEp588kn8+uuvePnllw3J1MnJCQqFwqgH6NSpU9ixY8e1Y2rnBFfndlYtHdpdazW6u7sbLc0lKJVKhdDQUOTn5xvKdDod8vPzW2yoXLx40SQJ6VuqlvQkWdSCCggIwMaNGxEfH4/+/fsbzSSxf/9+w7BgAAgODkZiYiJWrVqF8+fPY9iwYTh48CDWrVuHuLg43HPPPZZU3apx48bhpZdewsMPP4y//e1vuHjxIlasWIHbb7/d6ILgvHnzsHfvXowaNQo9e/ZEVVUVli9fju7du7d6F/yiRYswcuRIREZGYuLEiYZh5h4eHjZpDrdEqVTilVdeueF6Dz74IObNm4ekpCRERUXh8OHDePfdd036yAMCAtCxY0dkZ2fDzc0NHTp0QEREhMXXc/bs2YPly5cjPT3dMOx9zZo1GD58OGbNmoU33njDov0BwJQpU7By5UpMmDABxcXF8Pf3x9atW7Fv3z5kZWWZPTinJaWlpViwYIFJ+Z133tnqj6XOnTsjKSkJy5cvx3fffWfxzai21tZ4zB1mrqf/PBQVFZmsf/vtt8PLywtFRUUYOHCg0Y3zL774Iv7973/jwQcfNNwmUl9fj8OHD2Pr1q04deqUUav+enPmzMHu3bsRHR2NadOmQavV4q233sKAAQPaPD1XaGgoAODll1/GuHHj4OzsjNGjRyMgIAALFixAWloaTp06hbi4OLi5ueHkyZPYtm0bpkyZghdeeAHOzs5YsGABpk6dinvvvRfx8fE4efIk1qxZY9E1qDNnzmDDhg0ArrWajh49ii1btqCiogLPP/88pk6dalh31KhRyMzMxJ///GeMHz8eVVVVWLZsGXr37o1vv/22TefBWqmpqUhMTERYWBjCw8ORlZVl6EUBgISEBHTr1g0ZGRkAgNGjRyMzMxN33nmnoYtv1qxZGD16tGWXUiwa8/c/P/zwg5g8ebLw9/cXKpVKuLm5iejoaLF06VJx+fJlw3pXrlwRc+fOFb169RLOzs7Cz89PpKWlGa0jRMvDMJsOb25pmLkQQuzevVsMGDBAqFQq0bdvX7FhwwaTYeb5+flizJgxomvXrkKlUomuXbuKxx9/3GhIbHPDzIUQ4pNPPhHR0dGiffv2wt3dXYwePVocPXrUaB19fU2HseuH+DYd6trU9cPMW9LSMPPnn39edOnSRbRv315ER0eLoqKiZoeHb9++XQQFBRmG7uqPc9iwYeKOO+5ots7r91NbWyt69uwp/vSnP4krV64YrTdjxgyhVCpFUVFRq8fQ0r93ZWWlSEpKEl5eXkKlUomBAwea/Du09jfQWn1oYRj2xIkThRCtn/sff/xRODk5NTuctzW2GGZubjwAjIYBN2XJMHO9rl27CgBi1apVJu899NBDAoCYNm2ayXsXLlwQaWlpIjAwUKhUKuHl5SWioqLE4sWLRWNjo1HM1w8zF+LaZ/TOO+8UKpVKBAQEiHfeeUc8//zzQqPRGK3X0vH27NnT5N9p/vz5olu3bkKpVJp8Dt977z0xZMgQ0aFDB9GhQwfRr18/kZycLI4dO2a0j+XLl4tevXoJtVotwsLCxN69e5v9fDXn+r8/hUIh3N3dxR133CEmT54sPv/882a3Wb16tejTp49Qq9WiX79+Ys2aNWLmzJkCgHivU1+R5xlk1fJep74CgKipqblh/HpLly4VPXr0ECqVSoSHh4sDBw4Y3hs2bJjReb9y5YqYM2eOCAgIEBqNRvj5+YlnnnlG/Pbbb2bXJ4QQCiEsaG8REd1icXFx+Pbbb3H8+HF7h2JXtbW18PDwwDavfuigtG5AV71Oi4fPXevyNXdksT3I7nEbRPTH1XQW/+PHj2Pnzp18rIWDurlTCRMRWaB3796YMGECevfujdOnT2PFihVQqVQtDlt3RApnBRRK60bkKnS2G9F7MzFBEZFk/PnPf8amTZtQUVEBtVqNyMhILFy4sMUbah2Rsp0CSisTlJIJiojIMmvWrLF3CCQhTFBERDLCLj4iIpIkpZMCSicru/i08khQHMVHRESSxBYUEZGMKJwUUFjZglKALSiHY+kjkR3R3r17MXr0aHTt2hUKhQIffPCBvUOSpIyMDAwePBhubm7w9vZGXFxcsxMWE7BixQrDs6Dc3d0RGRlp1fx9Uqfv4rN2kQMmKBvRPxI5PT0dhw4dQnBwMGJjY1t8/o2jqq+vR3BwMJYtW2bvUCTts88+Q3JyMg4cOICPP/4YV65cwf3334/6+np7hyY53bt3x2uvvYbi4mJ8+eWXuPfeezFmzBi7zVt3symUCpsscsCpjmwkIiICgwcPxltvvQXg2my/fn5+ePbZZzFz5kw7RydNCoUC27ZtM3mgH5mqrq6Gt7c3PvvsM9x99932DkfyPD09sWjRIkycONHeodiMfqqjXQOD0cHKZ9fVa7WIPfw1pzpyBG15JDKRJfSPyfD09LRzJNKm1WqxefNm1NfX3/CZdXKlcFLaZJEDDpKwgbY8EpnIXDqdDtOnT0d0dDQGDBhg73Ak6fDhw4iMjMTly5fh6uqKbdu2GT088o/EJsPMZTJIggmKSOKSk5Nx5MgRFBYW2jsUyerbty9KSkpQU1ODrVu3IjExEZ999tkfNkk5CiYoG2jLI5GJzJGSkoIdO3Zg79696N69u73DkSyVSoXAwEAA1x5S+MUXX2DJkiVYuXKlnSOzPYXCcWaSkEdHpMS15ZHIRK0RQiAlJQXbtm3Dnj17LH7qsaPT6XRoaGiwdxg3hcLJ+qHmCuvGWNwybEHZyI0eiUzX1NXVobS01PD65MmTKCkpgaenJ3r06GHHyKQlOTkZGzduxPbt2+Hm5oaKigoAgIeHB9q3b2/n6KQlLS0NI0eORI8ePXDhwgVs3LgRBQUF2LVrl71DIysxQdlIfHw8qqurMXv2bFRUVCAkJAR5eXkmAycc3Zdffol77rnH8Do1NRUAkJiYiLVr19opKulZsWIFAJg8qG/NmjWYMGHCrQ9IwqqqqpCQkICzZ8/Cw8MDgwYNwq5duzBixAh7h3ZT2GQmCSGPLj7eB0VEJAP6+6AKhoTDtZ11bYu6q1cxvPAg74MiIiJqC3bxERHJiC2mKpLLVEdMUEREMmKTG3Vlcg2KXXxERCRJbEEREckIu/iIiEiSFAolFErrOr8UCnl0njFBERHJiCO1oOSRRmWkoaEBc+bM+cNOs2IrPE/m47kyD8/THw9v1LUx/c10Ur8Bzt54nszHc2WeP/p50h/fgZFD4eps5Y26V67iro/+K/lzxS4+IiIZYRcfERGRncm6BaXT6fDzzz/Dzc0NCoU0fhHU1tYa/Zeax/NkPp4r80jxPAkhcOHCBXTt2hVKK0fe6SmUNhjFZ6NYbjZZJ6iff/4Zfn5+9g6jWVKNS2p4nszHc2UeKZ6n8vJymz1w0pG6+GSdoNzc3AAAofdtgVM7FztHI23unTvaOwTZqP3lvL1DkIWBQwbYOwTJa7x8AetfHWD4riLLyDpB6bv1nNq5oJ1zBztHI23tVK72DkE22jlfsXcIsqDSSHf0l9TY8hIEW1BERCRJjpSg5HGljIiIHA4TFBGRjFxrQSmtXCxvQS1btgz+/v7QaDSIiIjAwYMHW1x3+PDhUCgUJsuoUaMsqpMJiohIRhRKheGZUG1dLE1Qubm5SE1NRXp6Og4dOoTg4GDExsaiqqqq2fXff/99nD171rAcOXIETk5OePTRRy2qlwmKiIhalZmZicmTJyMpKQlBQUHIzs6Gi4sLcnJyml3f09MTvr6+huXjjz+Gi4uLxQmKgySIiGTEloMkmt7UrFaroVarjcoaGxtRXFyMtLQ0Q5lSqURMTAyKiorMqm/16tUYN24cOnSwbLQ1W1BERDJi/fWn32ei8PPzg4eHh2HJyMgwqe/cuXPQarXw8fExKvfx8UFFRcUN4z148CCOHDmCSZMmWXysbEEREcmILVtQ5eXlRrOZN2092cLq1asxcOBAhIeHW7wtExQRkYNyd3e/4eM2vLy84OTkhMrKSqPyyspK+Pr6trptfX09Nm/ejHnz5rUpPnbxERHJiL4FZe1iLpVKhdDQUOTn5xvKdDod8vPzERkZ2eq2W7ZsQUNDA5544ok2HStbUEREMmKP2cxTU1ORmJiIsLAwhIeHIysrC/X19UhKSgIAJCQkoFu3bibXsFavXo24uDh07ty5TXEyQRERUavi4+NRXV2N2bNno6KiAiEhIcjLyzMMnCgrKzN5nMixY8dQWFiI3bt3t7leJigiIhmx11x8KSkpSElJafa9goICk7K+fftCCGFxPddjgiIikhFHemChPKIkIiKHwxYUEZGcKBTXFmv3IQNMUEREMqJQ2OAalEwSFLv4iIhIktiCIiKSEUcaJMEERUQkI3zkOxERkZ2xBUVEJCPs4iMiIklSKK3volPIIz8xQRERyQmvQREREdkZW1BERHKiVF5brN2HDDBBERHJiEKhsHomCM4kQUREZAW2oIiIZITDzImISJI4iu8WWrZsGfz9/aHRaBAREYGDBw/aOyQiIpIAuyao3NxcpKamIj09HYcOHUJwcDBiY2NRVVVlz7CIiKRLofx9JF9bF5ncqWvXKDMzMzF58mQkJSUhKCgI2dnZcHFxQU5Ojj3DIiKSrv918VmzgF18rWtsbERxcTFiYmJ+D0apRExMDIqKiprdpqGhAbW1tUYLERH9MdktQZ07dw5arRY+Pj5G5T4+PqioqGh2m4yMDHh4eBgWPz+/WxEqEZFkKBRKmyxyII8o/yctLQ01NTWGpby83N4hERHdWvouOmsXGbDbMHMvLy84OTmhsrLSqLyyshK+vr7NbqNWq6FWq29FeEREZGd2a0GpVCqEhoYiPz/fUKbT6ZCfn4/IyEh7hUVEJGn6G3WtXeTArjfqpqamIjExEWFhYQgPD0dWVhbq6+uRlJRkz7CIiCTLkW7UtWuCio+PR3V1NWbPno2KigqEhIQgLy/PZOAEERH9j0Jh/X1MMpks1u5THaWkpCAlJcXeYRARkcTYPUEREZH52MVHRETS5EAPLJRHlERE5HDYgiIikhFHeqIuExQRkZwobNDFx6mOiIiI2o4tKCIiGXGkUXxsQRERyYlCaZvFQpY+/fz8+fNITk5Gly5doFarcfvtt2Pnzp0W1ckWFBERtUr/9PPs7GxEREQgKysLsbGxOHbsGLy9vU3Wb2xsxIgRI+Dt7Y2tW7eiW7duOH36NDp27GhRvUxQRERyYovHZVi4/fVPPweA7OxsfPjhh8jJycHMmTNN1s/JycGvv/6K/fv3w9nZGQDg7+9veZgWb0FERHZjywcWNn1CeUNDg0l9bXn6+b///W9ERkYiOTkZPj4+GDBgABYuXAitVmvRsTJBERHJiQ0fWOjn52f0lPKMjAyT6try9PMTJ05g69at0Gq12LlzJ2bNmoV//vOfWLBggUWHyi4+IiIHVV5eDnd3d8NrWz0QVqfTwdvbG6tWrYKTkxNCQ0Nx5swZLFq0COnp6WbvhwmKiEhGbPHAQf327u7uRgmqOW15+nmXLl3g7OwMJycnQ1n//v1RUVGBxsZGqFQqs+JkFx8RkZwoFLZZzNSWp59HR0ejtLQUOp3OUPbDDz+gS5cuZicngAmKiIhuIDU1FW+//TbWrVuH7777DtOmTTN6+nlCQgLS0tIM60+bNg2//vornnvuOfzwww/48MMPsXDhQiQnJ1tUL7v4iIjkRKmwweM2LBtmfqOnn5eVlUF5XUx+fn7YtWsXZsyYgUGDBqFbt2547rnn8NJLL1lULxMUEZGcWNhF1+I+LNTa088LCgpMyiIjI3HgwAGL67keu/iIiEiS2IIiIpIRW47ikzomKCIiOWnjZK8m+5ABeURJREQOhy0oIiI5Udhgslg+8p2IiGzt+slerdmHHMgjSiIicjhsQRERyYkdngdlL0xQRERy4kCj+JigiIjkxE4zSdiDPNIoERE5nD9EC+qlVwajg2vrzzRxdPNmWTcnFlFTYX/ysHcIknep/ia0VJRKG0wWK4+2yR8iQREROQwHugYljyiJiMjhsAVFRCQnHGZORESSpFDYoItPHgmKXXxERCRJbEEREcmJA90HxQRFRCQnDjTMXB5REhGRw2ELiohITtjFR0REksQbdYmIiOyLLSgiIjlR2GCQhExaUExQRERywmtQREQkSbwGRUREZF9sQRERyQm7+IiISJI4kwQREZF9sQVFRCQjQqGAsLKLztrtbxUmKCIiOeHzoIiIiOyLLSgiIjlxoPugmKCIiGTEka5BySONEhGRw2GCIiKSE30Xn7WLhZYtWwZ/f39oNBpERETg4MGDLa67du1aKBQKo0Wj0VhcJxMUEZGc6GeSsHaxQG5uLlJTU5Geno5Dhw4hODgYsbGxqKqqanEbd3d3nD171rCcPn3a4kNlgiIiolZlZmZi8uTJSEpKQlBQELKzs+Hi4oKcnJwWt1EoFPD19TUsPj4+FtfLBEVEJCf6qY6sXQDU1tYaLQ0NDSbVNTY2ori4GDExMdeFoERMTAyKiopaDLOurg49e/aEn58fxowZg2+//dbyQ7V4CyIishv9KD5rFwDw8/ODh4eHYcnIyDCp79y5c9BqtSYtIB8fH1RUVDQbY9++fZGTk4Pt27djw4YN0Ol0iIqKwk8//WTRsXKYORGRnNjwPqjy8nK4u7sbitVqtXX7/Z/IyEhERkYaXkdFRaF///5YuXIl5s+fb/Z+mKCIiByUu7u7UYJqjpeXF5ycnFBZWWlUXllZCV9fX7PqcXZ2xp133onS0lKL4mMXHxGRjAiF0iaLuVQqFUJDQ5Gfn28o0+l0yM/PN2oltUar1eLw4cPo0qWLRcfKFhQRkZzY4YGFqampSExMRFhYGMLDw5GVlYX6+nokJSUBABISEtCtWzfDNax58+bhrrvuQmBgIM6fP49Fixbh9OnTmDRpkkX12jVB7d27F4sWLUJxcTHOnj2Lbdu2IS4uzp4hERFRE/Hx8aiursbs2bNRUVGBkJAQ5OXlGQZOlJWVQXndQxB/++03TJ48GRUVFejUqRNCQ0Oxf/9+BAUFWVSvXRNUfX09goOD8dRTT2Hs2LH2DIWISBYELOuia2kflkpJSUFKSkqz7xUUFBi9fvPNN/Hmm2+2JTQjdk1QI0eOxMiRI+0ZAhGRvNihi89eZHUNqqGhwehGstraWjtGQ0REN5OsRvFlZGQY3VTm5+dn75CIiG4t/RN1rVrk0YKSVYJKS0tDTU2NYSkvL7d3SEREt5QtZ5KQOll18anVapvd6UxERNImqwRFROTw+Mj3W6Ours5o6ouTJ0+ipKQEnp6e6NGjhx0jIyKSJgEFBKx85LuV298qdk1QX375Je655x7D69TUVABAYmIi1q5da6eoiIhICuyaoIYPHw4hhD1DICKSFUvn0mtpH3LAa1BERHLCa1BERCRFthgmLpdh5vJIo0RE5HDYgiIikhFegyIiImlyoMli5ZFGiYjI4bAFRUQkJzbo4uMoPiIisjlHmklCHmmUiIgcDltQREQywlF8REQkTQrYYBSfTSK56eSRRomIyOGwBUVEJCMCSggr2xbWbn+rMEEREckI5+IjIiKyM7agiIhkhKP4iIhIkhzpRl0mKCIiGXGkFpQ8oiQiIofDFhQRkYw40ig+JigiIhlxpGtQ7OIjIiJJYguKiEhGHGmQBBMUEZGMsIuPiIjIztiCIiKSEQEbdPHJpG0ijyiJiAjA71181i6WWrZsGfz9/aHRaBAREYGDBw+atd3mzZuhUCgQFxdncZ1MUERE1Krc3FykpqYiPT0dhw4dQnBwMGJjY1FVVdXqdqdOncILL7yAoUOHtqleJigiIhm5dqOu0srFshZUZmYmJk+ejKSkJAQFBSE7OxsuLi7IyclpcRutVou//vWvmDt3Lnr37t2mY2WCIiKSkVvdxdfY2Iji4mLExMQYypRKJWJiYlBUVNTidvPmzYO3tzcmTpzY5mP9QwyS+ObHdtC4/CEO5aZp58zzY66rV67aOwRZ8L6/n71DkLyLQmvvEFpVW1tr9FqtVkOtVhuVnTt3DlqtFj4+PkblPj4++P7775vdb2FhIVavXo2SkhKr4mMLiohIRvRz8Vm7AICfnx88PDwMS0ZGhtXxXbhwAU8++STefvtteHl5WbUv/qwmIpIRIRQQwsobdf+3fXl5Odzd3Q3lTVtPAODl5QUnJydUVlYalVdWVsLX19dk/R9//BGnTp3C6NGjDWU6nQ4A0K5dOxw7dgwBAQFmxckERUQkK0ob3Md0bXt3d3ejBNUclUqF0NBQ5OfnG4aK63Q65OfnIyUlxWT9fv364fDhw0Zlr7zyCi5cuIAlS5bAz8/P7CiZoIiIqFWpqalITExEWFgYwsPDkZWVhfr6eiQlJQEAEhIS0K1bN2RkZECj0WDAgAFG23fs2BEATMpvhAmKiEhG7DEXX3x8PKqrqzF79mxUVFQgJCQEeXl5hoETZWVlUCptP6SBCYqISEbsNVlsSkpKs116AFBQUNDqtmvXrrW4PoCj+IiISKLYgiIikhFHetwGExQRkYw4UoJiFx8REUkSW1BERDJiyxt1pY4JiohIRtjFR0REZGdsQRERyYgjtaCYoIiIZIQJioiIJEnABoMkZJKgeA2KiIgkiS0oIiIZ0UEBnZUtIGu3v1WYoIiIZMSRrkGxi4+IiCSJLSgiIhnhTBJERCRJAtZ30QnbhHLTsYuPiIgkiS0oIiIZYRcfERFJEkfxERER2RlbUEREMsIuPiIikiQBQGeDfcgBu/iIiEiS2IIiIpIRdvEREZEkOdIoPiYoIiIZcaQWFK9BERGRJNktQWVkZGDw4MFwc3ODt7c34uLicOzYMXuFQ0QkC/ouPmsXObBbgvrss8+QnJyMAwcO4OOPP8aVK1dw//33o76+3l4hERFJnk7YZpEDu12DysvLM3q9du1aeHt7o7i4GHfffbedoiIiIqmQzCCJmpoaAICnp2eL6zQ0NKChocHwura29qbHRUQkJY40ik8SgyR0Oh2mT5+O6OhoDBgwoMX1MjIy4OHhYVj8/PxuYZRERPanH8Vn7SIHkkhQycnJOHLkCDZv3tzqemlpaaipqTEs5eXltyhCIiK61ezexZeSkoIdO3Zg79696N69e6vrqtVqqNXqWxQZEZH0CHFtsXYfcmC3BCWEwLPPPott27ahoKAAvXr1slcoRESyoYMCOiuvIVm7/a1itwSVnJyMjRs3Yvv27XBzc0NFRQUAwMPDA+3bt7dXWEREJBF2uwa1YsUK1NTUYPjw4ejSpYthyc3NtVdIRESSZ69BEsuWLYO/vz80Gg0iIiJw8ODBFtd9//33ERYWho4dO6JDhw4ICQnB+vXrLa7Trl18RERkGXtcg8rNzUVqaiqys7MRERGBrKwsxMbG4tixY/D29jZZ39PTEy+//DL69esHlUqFHTt2ICkpCd7e3oiNjTW7XkmM4iMiIunKzMzE5MmTkZSUhKCgIGRnZ8PFxQU5OTnNrj98+HA8/PDD6N+/PwICAvDcc89h0KBBKCwstKheJigiIhmx5Vx8tbW1Rsv1EyHoNTY2ori4GDExMYYypVKJmJgYFBUV3TheIZCfn49jx45ZPEsQExQRkYzYci4+Pz8/o8kPMjIyTOo7d+4ctFotfHx8jMp9fHwMg9uaU1NTA1dXV6hUKowaNQpLly7FiBEjLDpWu98HRUREFrDFTBD/2768vBzu7u6GYlveZ+rm5oaSkhLU1dUhPz8fqamp6N27N4YPH272PpigiIgclLu7u1GCao6XlxecnJxQWVlpVF5ZWQlfX98Wt1MqlQgMDAQAhISE4LvvvkNGRoZFCYpdfEREMqIfxWftYi6VSoXQ0FDk5+cbynQ6HfLz8xEZGWn2fnQ6XbPXuFrDFhQRkYzYYyaJ1NRUJCYmIiwsDOHh4cjKykJ9fT2SkpIAAAkJCejWrZvhGlZGRgbCwsIQEBCAhoYG7Ny5E+vXr8eKFSssqpcJioiIWhUfH4/q6mrMnj0bFRUVCAkJQV5enmHgRFlZGZTK3zvk6uvr8cwzz+Cnn35C+/bt0a9fP2zYsAHx8fEW1csERUQkI/aaLDYlJQUpKSnNvldQUGD0esGCBViwYEEbIjPGBEVEJCO2eJ4TnwdFRERkBbagiIhk5Pobba3ZhxwwQRERyYgjPbCQXXxERCRJbEEREcnI9ZO9WrMPOWCCIiKSER1scA3KJpHcfOziIyIiSWILiohIRhxpkAQTFBGRjDBBERGRJOmEAjorZ4KwdvtbhdegiIhIktiCIiKSEXbxERGRJDlSgmIXHxERSRJbUEREMiJsMFmsXFpQTFBERDLC50ERERHZGVtQREQy4kiDJJigiIhkhA8slJnCXUfRTuVq7zAk7eqVq/YOgf5gMkausncIknf1Sj2wa5S9w5CtP0SCIiJyFOziIyIiSXKkBMVRfEREJElsQRERyQgHSRARkSQ5UhcfExQRkYzodNcWa/chB7wGRUREksQWFBGRjLCLj4iIJMmREhS7+IiISJLYgiIikhEdbDDM3CaR3HxMUEREMiKEgLCyj87a7W8VdvEREZEksQVFRCQjHCRBRESSJHS/36zb1kW04SLUsmXL4O/vD41Gg4iICBw8eLDFdd9++20MHToUnTp1QqdOnRATE9Pq+i1hgiIiolbl5uYiNTUV6enpOHToEIKDgxEbG4uqqqpm1y8oKMDjjz+OTz/9FEVFRfDz88P999+PM2fOWFQvExQRkYzou/isXSyRmZmJyZMnIykpCUFBQcjOzoaLiwtycnKaXf/dd9/FM888g5CQEPTr1w/vvPMOdDod8vPzLaqXCYqISEb0s5lbu5irsbERxcXFiImJMZQplUrExMSgqKjIrH1cvHgRV65cgaenp0XHykESREQOqra21ui1Wq2GWq02Kjt37hy0Wi18fHyMyn18fPD999+bVc9LL72Erl27GiU5c7AFRUQkI7bs4vPz84OHh4dhycjIsHm8r732GjZv3oxt27ZBo9FYtC1bUEREMiJ0AsLKqST025eXl8Pd3d1Q3rT1BABeXl5wcnJCZWWlUXllZSV8fX1brWfx4sV47bXX8Mknn2DQoEEWx8kWFBGRjNjyGpS7u7vR0lyCUqlUCA0NNRrgoB/wEBkZ2WKcb7zxBubPn4+8vDyEhYW16VjZgiIiolalpqYiMTERYWFhCA8PR1ZWFurr65GUlAQASEhIQLdu3QxdhK+//jpmz56NjRs3wt/fHxUVFQAAV1dXuLq6ml0vExQRkYzYYyaJ+Ph4VFdXY/bs2aioqEBISAjy8vIMAyfKysqgVP7eIbdixQo0NjbikUceMdpPeno65syZY3a9TFBERDKi0wnorLwG1ZbtU1JSkJKS0ux7BQUFRq9PnTrVhqhM8RoUERFJEltQREQy4kiTxTJBERHJiCMlKHbxERGRJLEFRUQkIzohoLOyCWTt9rcKExQRkYyINj7Pqek+5IBdfEREJElsQRERyYiAgLCyi05AHl18dmtBrVixAoMGDTLMARUZGYmPPvrIXuEQEcmCvR75bg92a0F1794dr732Gvr06QMhBNatW4cxY8bgq6++wh133GGvsIiIJE0IG7SgOEiidaNHjzZ6/eqrr2LFihU4cOAAExQREUnjGpRWq8WWLVtQX1/f6vTtDQ0NaGhoMLxu+jRIIqI/Oksf2d7SPuTArgnq8OHDiIyMxOXLl+Hq6opt27YhKCioxfUzMjIwd+7cWxghEZG02PKBhVJn12Hmffv2RUlJCT7//HNMmzYNiYmJOHr0aIvrp6WloaamxrCUl5ffwmiJiOhWsmsLSqVSITAwEAAQGhqKL774AkuWLMHKlSubXV+tVjf7xEciIkfhSHPxSeIalJ5OpzO6xkRERMbs9Twoe7BbgkpLS8PIkSPRo0cPXLhwARs3bkRBQQF27dplr5CIiEhC7JagqqqqkJCQgLNnz8LDwwODBg3Crl27MGLECHuFREQkebwP6hZYvXq1vaomIpItThZLRERkZ5IaJEFERK3j86CIiEiSHOkaFLv4iIhIktiCIiKSEd4HRUREksSZJIiISJKEsMFksTLJULwGRUREksQWFBGRjAgbDDOXSwuKCYqISEb4PCgiIiI7YwuKiEhGHKkFxQRFRCQjOnFtsXYfcsAuPiIikiS2oIiIZIRdfEREJEmcLJaIiMjOmKCIiGREp/t9wti2L5bXu2zZMvj7+0Oj0SAiIgIHDx5scd1vv/0Wf/nLX+Dv7w+FQoGsrKw2HSsTFBGRjOi7+KxdLJGbm4vU1FSkp6fj0KFDCA4ORmxsLKqqqppd/+LFi+jduzdee+01+Pr6tvlYmaCIiKhVmZmZmDx5MpKSkhAUFITs7Gy4uLggJyen2fUHDx6MRYsWYdy4cVCr1W2ul4MkiIhkxJaj+Gpra43K1Wq1SUJpbGxEcXEx0tLSDGVKpRIxMTEoKiqyKo4bYQuKiEhG9AnK2gUA/Pz84OHhYVgyMjJM6jt37hy0Wi18fHyMyn18fFBRUXFTj5UtKCIiGdHB+tnMdbi2fXl5Odzd3Q3l1nTH3QxMUEREDsrd3d0oQTXHy8sLTk5OqKysNCqvrKy0agCEOdjFR0QkI7bs4jOHSqVCaGgo8vPzDWU6nQ75+fmIjIy8GYdowBYUEZGM2GMmidTUVCQmJiIsLAzh4eHIyspCfX09kpKSAAAJCQno1q2b4RpWY2Mjjh49avj/M2fOoKSkBK6urggMDDS7XiYoIiJqVXx8PKqrqzF79mxUVFQgJCQEeXl5hoETZWVlUCp/75D7+eefceeddxpeL168GIsXL8awYcNQUFBgdr1MUEREMiL+NxuEtfuwVEpKClJSUpp9r2nS8ff3t8l8f0xQREQywtnMZUKfoa9eqbdzJNLHc0R062mvXgQgn9nDpUbWCerChQsAgM+2xto5EiKill24cAEeHh422ZcjPW5D1gmqa9euKC8vh5ubGxQKhb3DAXBt6hA/Pz+TG+DIGM+T+XiuzCPF8ySEwIULF9C1a1fb7VOng2jLdORN9iEHsk5QSqUS3bt3t3cYzTLnBjjiebIEz5V5pHaebNVyckSyTlBERI5GZ4NRfNZuf6swQRERyYgjXYPiVEc2plarkZ6eLrlJF6WG58l8PFfm4Xn641EIuaRSIiIHVltbCw8PD4yZ9jWc1W5W7etKwwVsXxGMmpoaSV2va4pdfEREMsIbdYmISJJ00EEnrBsmroM8hpnzGhQREUkSW1BERDIidNZ30VnZALtlmKCIiGTEka5BsYuPiIgkiS0oIiIZcaQbdZmgiIhkRKfTQWflZK/Wbn+rsIuPiIgkiS0oIiIZcaRBEkxQREQyIoQOwspx4tZuf6uwi4+IiCSJLSgiIhlhFx8REUmTDRIUZJKg2MVHRESSxBYUEZGM6IQNZjOXySAJJigiIhnhNSgiIpIkIXQQVs4EwWHmREREVmALiohIRtjFR0REksSZJIiIiOyMLSgiIhnR6QCdlV10MnnaBhMUEZGcCJ0NRvHJJEOxi4+IiCSJLSgiIhlxpFF8bEEREcmIfhSftYulli1bBn9/f2g0GkRERODgwYOtrr9lyxb069cPGo0GAwcOxM6dOy2ukwmKiIhalZubi9TUVKSnp+PQoUMIDg5GbGwsqqqqml1///79ePzxxzFx4kR89dVXiIuLQ1xcHI4cOWJRvQohhDzaekREDqy2thYeHh64a+RHaOfcwap9Xb1SjwMfjURNTQ3c3d1vuH5ERAQGDx6Mt956CwCg0+ng5+eHZ599FjNnzjRZPz4+HvX19dixY4eh7K677kJISAiys7PNjpMtKCIiGdGP4rN2MVdjYyOKi4sRExNjKFMqlYiJiUFRUVGz2xQVFRmtDwCxsbEtrt8SDpIgIpIR7dV6m+2jtrbWqFytVkOtVhuVnTt3DlqtFj4+PkblPj4++P7775vdf0VFRbPrV1RUWBQnExQRkQyoVCr4+vriy/zHbLI/V1dX+Pn5GZWlp6djzpw5Ntm/LTBBERHJgEajwcmTJ9HY2GiT/QkhoFAojMqatp4AwMvLC05OTqisrDQqr6yshK+vb7P79vX1tWj9ljBBERHJhEajgUajuaV1qlQqhIaGIj8/H3FxcQCuDZLIz89HSkpKs9tERkYiPz8f06dPN5R9/PHHiIyMtKhuJigiImpVamoqEhMTERYWhvDwcGRlZaG+vh5JSUkAgISEBHTr1g0ZGRkAgOeeew7Dhg3DP//5T4waNQqbN2/Gl19+iVWrVllULxMUERG1Kj4+HtXV1Zg9ezYqKioQEhKCvLw8w0CIsrIyKJW/DwqPiorCxo0b8corr+Af//gH+vTpgw8++AADBgywqF7eB0VERJLE+6CIiEiSmKCIiEiSmKCIiEiSmKCIiEiSmKCIiEiSmKCIiEiSmKCIiEiSmKCIiEiSmKCIiEiSmKCIiEiSmKCIiEiSmKCIiEiS/h95PeI1nIgHnwAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "acc1 = accuracy_score(y_test1, rf_model1_pred_weight)\n",
        "pre1 = precision_score(y_test1, rf_model1_pred_weight, average= 'macro')\n",
        "rec1 = recall_score(y_test1, rf_model1_pred_weight, average='macro')\n",
        "f11 = f1_score(y_test1, rf_model1_pred_weight, average= 'macro')\n",
        "\n",
        "print(acc1, pre1, rec1, f11)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QBKBPW8G__FP",
        "outputId": "1c8bb7c0-bc1a-4fa6-be8e-f0dd8b71ffa3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.566 0.4920334793568545 0.5660000000000001 0.4840707493349233\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "rf_model1_proba_weight = lr_rf_model_final_weight.predict_proba(sel_x_test_final_weight)"
      ],
      "metadata": {
        "id": "pe-jtPDXAAJR"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "Y_binary1 = label_binarize(y_test1, classes=[0,1,2,3])\n",
        "\n",
        "plt.figure()\n",
        "plt.clf()\n",
        "plt.plot([0,1], [0,1], 'k--')\n",
        "plt.axis([0,1,0,1])\n",
        "plt.xlabel('FPR')\n",
        "plt.ylabel('TPR')\n",
        "plt.title('Avg ROC')\n",
        "\n",
        "Fpr_mean = np.linspace(0,1,100)\n",
        "\n",
        "\n",
        "classifier1 = rf_model1_proba_weight\n",
        "labelval = 'LR RF'\n",
        "color = 'b'\n",
        "\n",
        "\n",
        "tpr_mean = []\n",
        "\n",
        "for j in range(0, np.size(Y_binary1, 1)):\n",
        "  fpr, tpr, T = roc_curve(Y_binary1[:,j], classifier1[:,j])\n",
        "  tpr_mean.append(np.interp(Fpr_mean, fpr, tpr))\n",
        "Tpr_mean = np.mean(tpr_mean, axis=0)\n",
        "mean_auc = auc(Fpr_mean, Tpr_mean)\n",
        "plt.plot(Fpr_mean, Tpr_mean, color=color, label= r'%s (AUC = %0.2f)' % (labelval,mean_auc))\n",
        "\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "HtgwzZ92AA5H",
        "outputId": "efc43c9a-12b3-46ac-b472-3db51861e86a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from pylab import *\n",
        "\n",
        "figure()\n",
        "gray()\n",
        "contour()"
      ],
      "metadata": {
        "id": "dkzRegb8GGH_"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}