Untitled8.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Untitled8.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyP4ZGkEf4D2aduQH4vQ25Ur",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/dewhitee/ml/blob/main/Untitled8.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pKbBMhv3I2DL"
      },
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn import linear_model\n",
        "from scipy.special import expit # logistical curve\n",
        "from sklearn.tree import DecisionTreeClassifier"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Hvg3xRmkL9BL"
      },
      "source": [
        "x = np.random.normal(size=100) \n",
        "# larger 0 = cat, less = not cat\n",
        "y = (x > 0).astype(np.float) # assigning classes\n",
        "x[x > 0] *= 5. # add shift to x's\n",
        "x += 0.3 * np.random.normal(size=100) # add noise\n",
        "# adding new axis to transform list to numpy array as we need numpy array to work with expit\n",
        "#print(x[:5])\n",
        "x = x[:, np.newaxis]\n",
        "#print(x)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "shPVQTxFMuOc",
        "outputId": "b98961a7-717c-4641-d920-7f53003a44d2"
      },
      "source": [
        "plt.scatter(x, y)\n",
        "plt.grid()\n",
        "#plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-Ldn060JNAsZ",
        "outputId": "2f86bc03-7feb-4ad5-a29d-d55705772caa"
      },
      "source": [
        "models = [DecisionTreeClassifier, linear_model.LogisticRegression]\n",
        "\n",
        "for i in range(len(models)):\n",
        "    clf = models[i]()\n",
        "    clf.fit(x, y)\n",
        "    print('Accuracy: ', clf.score(x, y))\n",
        "\n",
        "## machine learning - creating classificator\n",
        "#clf = linear_model.LogisticRegression()\n",
        "## training\n",
        "#clf.fit(x, y)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Accuracy:  1.0\n",
            "Accuracy:  0.97\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vmuNv8RgNhhW",
        "outputId": "0492d5c8-6c41-4b59-c355-960ff0b25191"
      },
      "source": [
        "print('Accuracy: ', clf.score(x, y))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Accuracy:  0.97\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "g7ZQFd39NmpN",
        "outputId": "18ea83b3-d011-41a2-9ce7-732985d77603"
      },
      "source": [
        "for i in range(len(y)):\n",
        "    if y[i] == 0: # if first class\n",
        "        plt.scatter(x[i], y[i], color='blue')\n",
        "    else: # if second class\n",
        "        plt.scatter(x[i], y[i], color='green')\n",
        "\n",
        "x_span = np.linspace(-4, 8, 200)\n",
        "logReg = (expit(x_span*clf.coef_ + clf.intercept_)).ravel()\n",
        "plt.plot(x_span, logReg, color='red', linewidth=3)\n",
        "plt.grid()\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAD4CAYAAAD8Zh1EAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAeW0lEQVR4nO3de5QcdZn/8fczM5mQmwkhOCKTZJIQkRi8MJFwEUxIxAALcZVVIosE0BxAPHgUFRdFD5izi66s/lai5Aes4EYjXoAsBBFxRsQLkgRISELCEDK5kCUYQmByncuzf1RPumfSt5l0d3VXfV7n9JlvVX27+pnqmc/UfKuqy9wdERGpfFVhFyAiIoWhQBcRiQgFuohIRCjQRUQiQoEuIhIRNWG98KhRo7yhoSGsl8/L7t27GTJkSNhl5KWSaoXKqle1Fodq7Z/ly5f/3d2PTrcstEBvaGhg2bJlYb18Xpqbm5k2bVrYZeSlkmqFyqpXtRaHau0fM2vNtExDLiIiEaFAFxGJCAW6iEhEKNBFRCJCgS4iEhE5z3Ixs7uAfwC2u/vkNMsN+D5wLrAHmOvuKwpdqOTn6oeuZuHyhXR6J9VWzbzGeSw4bwGLVi3ihsduYNOuTYwcNJJ9HfvY3b4bgKMGHcX3z/k+F5948cF+rbtaqbIqurzrkD4Ai1Yt4tqHr2XH3h0HXzu1T6Y6um3atYnqm6oPrj+dKquixmo40HUgYw1XPnglbQfaMq7DMM4adxYtr7Wwadcmxgwfw/wZ8w/WePvy2zPW0F33x4d8POP6U7dXtVXT6Z2MHT724Guk9un9+rnkel665cdybM715qu/dRfidVN/tnq/75Ws2NvUcn3aopmdCbQB92QI9HOBzxEE+lTg++4+NdcLT5kyxXXaYuE0Nzdz7+57+eGyHx6ybMa4Gfxly1/Y074n4/Nrq2u54n1XcPezd2fsV1tdy12z7wLgsvsvo72rPW2fM8acwWMvPXbIsqumXMWC8xZw9UNXM+GNCVy3/rp8vz1wGNAJw30AP5r5Par3t/Pl33yRrs5Oqh2qHKq7El/TTFvKj/kRNQOZ/NbJLH95+SEvY2l+Hf5l4leZdfqHD5n/6Iu/5bt//i77Ovb3XEfiNb546hcB+O5fvsv+9mSfI2oG8oVTv8DM8TMzfru/2/A7bv3LrT3Wnfq8TMu/etxXOWPqGRnXC0B1NXzwg1m7LFq1iHn/M6/Hz8LgAYNZeP7CggVQut+vRasWcfkDl3Og80CP+QOqBvBfH/mv0EK9EFlQqG1qZsvdfUraZfl8fK6ZNQAPZgj024Fmd/9ZYnodMM3dt2VbpwK9sJqbm5n5+Ew6vbPf6+jew8xm7PCxALTuyngqbNb1d9zYQc1NNdwy8Rb+Zc11nPAqjN8JY3fB2NfhrbvhqL0wci8ctQfesh8GdcARHUEwSwEMHQpvvpm1S8P3GtK+x2OHj2Xj5zcWpIx0v1+ZXrfQr91XhciCQm3TYgf6g8C/ufsTienHgK+4+yFpbWbzgHkAdXV1jYsXL877mwhDW1sbQ4cODbuMvLS1tbHuzXVhl5HVoDfamLa5nc4/PUp963besmkz1Z39/wMk/dMxaBBPLF2atc/ybYf+99Kt8ZjGgtSR7vcr2+sW8rX7qhBZUKhtOn369IyBXtIrRd19IbAQgj30ct/7rbQ99K+s+ErZ7aHXdsBH18JlT8PMl4IhkP5qr4K9NXCgtooDNcZ+Ouk06DLorEp8zTDdZenX6Wnm9y5x/OBx1B9dD9az85Nbn2Rvr+GW1PUeUTMQ4OCwSOp6j6gZyGmjT8v4vf55858PGcpJfV6m5ROHTGT0W+szrhegZtCgnD/Xc783N/Pe5JyNWZ+br3S/X5let9Cv3VeFyIJSbNNCBPpWYHTKdH1inpTYvMZ5RR9Dnz9jPpB7DP3xFx7j8qfhhj/C6Dcy17xhBKw9GjaOgNbh8PIw2DEYXhsUPHYNhL0DgiDvrO45jj/3/rl0dHVk2yRpDR4wmFPrT007zp/O3e/9Jp+a/alD5rekGRNNfY2F5y8EyDhuelqWcdOXsoy3nnbixRmX/+R9/8bocz6a1/eVzfwZ89Ouv/v9L5b5M+ZnHEMv9msXWym2aSECfQlwjZktJjgouivX+LkUR/dZJIdzlsvpY07P6ywXIPNZLvuP5+Wvn83bN+3sWaAZnHoqzJrFI7aHObtvYecRmXfZc53lApTkLJcxQ8akXd5dR66zXLr79OXMhtR1p3tepuUjd4zMut585Xr9YulefxTPcinJNnX3rA/gZ8A2oB3YAlwBXAlcmVhuwG3Ai8AqYEqudbo7jY2NXu6amprCLiFvZVFrV5f7t77lXl3tDslHXZ37zTe7t7Ye7FoW9eZJtRaHau0fYJlnyNWce+juPifHcgc+2/8/KRIJHR1w1VVwxx3JeUOGwDe+AZ/9LAweHF5tIjER2sfnSoR0dMAnPgG//nVy3gc+AD/+MUyYEFpZInGjS//l8H35yz3D/JJL4Pe/V5iLlJgCXQ7PnXfCf/xHcvraa4M98wEDQitJJK4U6NJ/Tz8djJt3+8d/hFtvhSr9WImEQb950j8HDsCll0J74lz097wH7rlHYS4SIv32Sf9873uwalXQHjQIfvGL4DNCRCQ0CnTpu23b4Oabk9Pf+hZMnBhePSICKNClP+bPh7bE1Znvehd87nPh1iMigAJd+mrTJli4MDn9ne/ojBaRMqFAl7659dbkgdDTToNZs8KtR0QOUqBL/nbu7Hlp/9e/fshHyopIeBTokr+77oLdwSc0MnkyfPjQ27KJSHgU6JIf955j59deq71zkTKjQJf8PP44rF8ftIcNgzlZP4RTREKgQJf83HNPsv3JTwYfjSsiZUWBLrnt2we/+lVyeu7c0EoRkcwU6JLbww/Drl1Be/x4mDo13HpEJC0FuuR2333J9pw5OhgqUqYU6JJdZycsXZqc/shHwqtFRLJSoEt2Tz4JO4K7r/O2t8FJJ4Vbj4hkpECX7B56KNk+7zx93rlIGdNvp2T34IPJ9nnnhVeHiOSkQJfMNm2ClSuDdm0tzJwZbj0ikpUCXTJLPRg6bVpwhaiIlC0FumT26KPJ9rnnhleHiORFgS7pucMf/5icPuus8GoRkbwo0CW99evh1VeD9pFHBreaE5GypkCX9FL3zk8/XacrilQA/ZZKeqmBfsYZ4dUhInlToEt6jz+ebJ95Znh1iEjeFOhyqC1bYOPGoD1okC73F6kQeQW6mc0ys3Vm1mJm16dZPsbMmszsaTNbaWY6x62SpQ63nHJKcFGRiJS9nIFuZtXAbcA5wCRgjplN6tXta8C97v4+4CJgQaELlRLS+LlIRcpnD/1koMXdN7j7AWAxMLtXHwfekmgPB14uXIlScn/7W7J9+unh1SEifWLunr2D2YXALHf/dGL6EmCqu1+T0ucY4LfAkcAQYKa7L0+zrnnAPIC6urrGxYsXF+r7KIq2tjaGDh0adhl5KVSt1t7OGeedR1V7OwBP3H8/HcOHH/Z6e4vjti0F1Voc5VTr9OnTl7v7lLQL3T3rA7gQuCNl+hLgB736fAH4YqJ9KrAGqMq23sbGRi93TU1NYZeQt4LVumKFe3CdqHtDQ2HWmUYst20JqNbiKKdagWWeIVfzGXLZCoxOma5PzEt1BXBv4g/EX4AjgFF5rFvKzfKUf6waG8OrQ0T6LJ9AfwqYaGbjzKyW4KDnkl59NgEzAMzsBIJAf7WQhUqJrFiRbCvQRSpKzkB39w7gGuARYC3B2SyrzewmM7sg0e2LwGfM7FngZ8DcxL8GUmlS99B1/rlIRanJp5O7LwWW9pp3Y0p7DaDTISpdezs8+2xyWnvoIhVFV4pK0po1sH9/0B4zBkbpMIhIJVGgS5LGz0UqmgJdknSGi0hFU6BL0tNPJ9s6ICpScRToEnCH555LTr/nPeHVIiL9okCXwJYt8MYbQXvECDjmmHDrEZE+U6BLYPXqZHvyZDALrxYR6RcFugRSh1t0Q2iRiqRAl0DvPXQRqTgKdAloD12k4inQBbq6gqtEu2kPXaQiKdAluCH0nj1B++ijg4eIVBwFuvQcbtHeuUjFUqBLzwOiGj8XqVgKdNEeukhEKNBFe+giEaFAj7uuLli/Pjk9aVJ4tYjIYVGgx93mzbB3b9AeNQpGjgy3HhHpNwV63K1bl2y/4x3h1SEih02BHnepwy3HHx9eHSJy2BTocZe6h65AF6loCvS4U6CLRIYCPe4U6CKRoUCPs717YdOmoF1VBRMmhFuPiBwWBXqcvfBCsj1uHNTWhleLiBw2BXqcabhFJFIU6HGmQBeJFAV6nOkcdJFIUaDHmfbQRSJFgR5X7j330HXZv0jFyyvQzWyWma0zsxYzuz5Dn4+b2RozW21mPy1smVJwr70Gr78etIcMgWOOCbceETlsNbk6mFk1cBvwIWAL8JSZLXH3NSl9JgJfBU53951m9tZiFSwF0tKSbE+YAGbh1SIiBZHPHvrJQIu7b3D3A8BiYHavPp8BbnP3nQDuvr2wZUrBvfhisq0LikQiIeceOnAssDllegswtVefdwCY2Z+AauCb7v6b3isys3nAPIC6ujqam5v7UXLptLW1lX2N3fpa69hHH2Vcor2ptpYNJf4+o7xtw6Rai6NianX3rA/gQuCOlOlLgB/06vMgcB8wABhH8AdgRLb1NjY2erlramoKu4S89bnWT33KPTg06n777UWpKZtIb9sQqdbiKKdagWWeIVfzGXLZCoxOma5PzEu1BVji7u3u/hKwHpjY3z8yUgK9x9BFpOLlE+hPARPNbJyZ1QIXAUt69bkfmAZgZqMIhmA2FLBOKbTUQD/uuPDqEJGCyRno7t4BXAM8AqwF7nX31WZ2k5ldkOj2CLDDzNYATcCX3H1HsYqWw/Tmm7A9cdy6thbq68OtR0QKIp+Dorj7UmBpr3k3prQd+ELiIeUu9QyXceOgujq8WkSkYHSlaBxpuEUkkhTocaRz0EUiSYEeR9pDF4kkBXocKdBFIkmBHkcachGJJAV63OzdC5sTn+RQVQUNDaGWIyKFo0CPm5deSrbHjtWNoUUiRIEeNxpuEYksBXrc6ICoSGQp0ONGH8olElkK9LhJHXLRHrpIpCjQ40ZDLiKRpUCPk/Z2aG1NTo8fH14tIlJwCvQ42bQJOjqC9tvfDoMHh1uPiBSUAj1ONNwiEmkK9DjROegikaZAjxPtoYtEmgI9TnQOukikKdDjROegi0SaAj0uuro0hi4ScQr0uNi6FfbvD9pHHQUjRoRbj4gUnAI9LlLHzydODK8OESkaBXpcvPBCsq3xc5FIUqDHhU5ZFIk8BXpcaMhFJPIU6HGhIReRyFOgx0HvUxYV6CKRpECPg23bYO/eoD1yZPAQkchRoMeBhltEYkGBHgc6ICoSCwr0ONApiyKxkFegm9ksM1tnZi1mdn2Wfh8zMzezKYUrUQ6bhlxEYiFnoJtZNXAbcA4wCZhjZpPS9BsGXAs8Wegi5TBpyEUkFvLZQz8ZaHH3De5+AFgMzE7T72bgFmBfAeuTw+WuIReRmKjJo8+xwOaU6S3A1NQOZnYSMNrdHzKzL2VakZnNA+YB1NXV0dzc3OeCS6mtra3sa+yWqdbav/+d0/bsAaB96FD+tHIlmJW4ukNFYduWI9VaHBVTq7tnfQAXAnekTF8C/CBlugpoBhoS083AlFzrbWxs9HLX1NQUdgl5y1jrH/7gHuynu7///SWtKZtIbNsypFqLo5xqBZZ5hlzNZ8hlKzA6Zbo+Ma/bMGAy0GxmG4FTgCU6MFomdEBUJDbyCfSngIlmNs7MaoGLgCXdC919l7uPcvcGd28A/gpc4O7LilKx9I0OiIrERs5Ad/cO4BrgEWAtcK+7rzazm8zsgmIXKIdJB0RFYiOfg6K4+1Jgaa95N2boO+3wy5KC0ZCLSGzoStEo633KooZcRCJNgR5lr7wCu3cH7eHDg5tDi0hkKdCjbP36ZPu448ri/HMRKR4FepQ9/3yy/c53hleHiJSEAj3KFOgisaJAjzIFukisKNCjbO3aZPuEE8KrQ0RKQoEeVXv2QGtr0K6q0jnoIjGgQI+qF14IzkMHGD8eBg4Mtx4RKToFelRp/FwkdhToUaVAF4kdBXpU6YCoSOwo0KNKe+gisaNAj6KuLli3Ljl9/PHh1SIiJaNAj6JNm2Bf4l7dRx+tD+USiQkFehSljp9ruEUkNhToUbRqVbI9eXJ4dYhISSnQo2jlymT7xBPDq0NESkqBHkWpgf7ud4dXh4iUlAI9ag4c6DmGriEXkdhQoEfNunXQ0RG0GxqCW8+JSCwo0KNG4+cisaVAjxqNn4vElgI9alJPWVSgi8SKAj1qtIcuElsK9CjZsQO2bg3aAwfqLkUiMaNAj5LU4ZZ3vQtqasKrRURKToEeJU8/nWxruEUkdhToUfK3vyXb739/eHWISCgU6FGiQBeJtbwC3cxmmdk6M2sxs+vTLP+Cma0xs5Vm9piZjS18qZJNza5dsGFDMFFbqyEXkRjKGehmVg3cBpwDTALmmNmkXt2eBqa4+7uBXwLfLnShkt1bUu9Q9N73Bme5iEis5LOHfjLQ4u4b3P0AsBiYndrB3ZvcfU9i8q9AfWHLlFyGpX4g18knh1eIiIQmn/PajgU2p0xvAaZm6X8F8HC6BWY2D5gHUFdXR3Nzc35VhqStra3sa+x2wnPPHWyvHTaMV8q87kratqq1OFRrEbh71gdwIXBHyvQlwA8y9P1ngj30gbnW29jY6OWuqakp7BLy09Xl+0eMcIfg8fzzYVeUU8VsW1etxaJa+wdY5hlyNZ899K3A6JTp+sS8HsxsJnAD8EF3338Yf2Okr1pbqX399aA9fDhMnBhuPSISinzG0J8CJprZODOrBS4ClqR2MLP3AbcDF7j79sKXKVn1Pl2xSmejisRRzt98d+8ArgEeAdYC97r7ajO7ycwuSHT7DjAU+IWZPWNmSzKsTorhj39MtnVAVCS28vqwD3dfCiztNe/GlPbMAtclfdHUlGxPmxZaGSISLv1vXuleeQVWrw7aAwbA6aeHW4+IhEaBXulS985POQUGDw6vFhEJlQK90qUG+llnhVeHiIROgV7pfv/7ZFuBLhJrCvRKtnkztLQA0FlbC1OzXcArIlGnQK9kKcMtu048UR/IJRJzCvRK9pvfHGy+/t73hliIiJQDBXqlam+Hh5OfgfbaKaeEWIyIlAMFeqV64gno/vyW0aNpmzAh3HpEJHQK9Er1wAPJ9vnng1l4tYhIWVCgV6KuLvjlL5PTs2dn7isisaFAr0R/+hNsTXyC8ahROv9cRAAFemVavDjZ/tjHoCavz1gTkYhToFeaffvgpz9NTs+ZE14tIlJWFOiV5r77kme3TJgAZ54Zbj0iUjYU6JXmRz9Kti+7TGe3iMhBCvRK8swz8PjjQbumBubODbUcESkvCvRKcuutyfaFF8Kxx4ZXi4iUHQV6pWhp6Xkw9Nprw6tFRMqSAr1S3HwzdHYG7RkzgrsTiYikUKBXghUr4Cc/SU5/4xvh1SIiZUuBXu66uuBznwP3YPr88+GMM8KtSUTKkgK93C1YAH/+c9CuqYFvfzvcekSkbCnQy9mqVfClLyWnr78e3vnO8OoRkbKmQC9XO3bARz8aXOoPMHkyfO1r4dYkImVNgV6Odu6E8847eANoBg2Cn/9c9wwVkaz0MX3lZvt2OPtsePbZ5Lz//m+YNCm8mkSkImgPvZw88QScfHLPMF+wIBh6ERHJQYFeDnbuhOuugw9+EFpbg3lVVfDjH8NVV4VamohUDg25hKm1Fe68E/7zP5MfiQtw5JFBmF9wQWiliUjlUaCXUns7PPccPPYYPPQQ/OEPyQuGuk2fDvfcA/X14dQoIhUrryEXM5tlZuvMrMXMrk+zfKCZ/Tyx/Ekzayh0oWG5+urgeh4zqK6GoUODdve8hgZYtCjRed8+ePllbv7Ec5xpj/NJ+yk32HzusE/zpE1lX+0wOOmk4Nzy5uYeYb6Od/BP3EtV0++w0fWY0afH8uWHzquuPrTGRYuC6aqqXrWLSMXLuYduZtXAbcCHgC3AU2a2xN3XpHS7Atjp7seZ2UXALcAnCl7txo3BlZLuyTDsbh/uI826xre+ytkrjuLDOFV0MbBrPwN372cg+zmicx8D2c/A1v0MumQfHZe/Ts2BvQB8PfHIx2/5EP+fz3A/H6GDAQXdXF1dwdfWVpg3L7i39N13w549PecDXHxxQV9aREKQz5DLyUCLu28AMLPFwGwgNdBnA99MtH8J/MDMzL33eMJh2r4dfvjDgq4ymzGJR04OHMhvnRsZy1O8n98wi4c5h228vf8F9sGePbBwYfIDG1Pn33CDAl0kCixX5prZhcAsd/90YvoSYKq7X5PS57lEny2J6RcTff7ea13zgHkAdXV1jYtT716fh2HPP09jGZ/10VVTw95Bw9g36C3sGzyM3cOO5I2Rx7Br5NvYOaqe7fXvYO+Q4UV57fr6NrZsGdrv5zc2FrCYPLS1tTF0aP/rLSXVWhyqtX+mT5++3N2npFtW0oOi7r4QWAgwZcoUnzZtWt9WcMIJwS5m6kAx9G2wOdMjzXoe+J/V/PjuE3GMLqoIBlsGso8jDmkPHz2cNa1DGD7A6HyzcNssX//+781cd920rH2qqw/dQwcYOzYYzSql5uZm+vz+h0S1FodqLbx8An0rMDpluj4xL12fLWZWAwwHdhSkwlR1dfDZzxZ8tZns2n0U9989LWe/wYPh2/8KWDAmXcJRobwNHgyXXtpzDL17/vz54dUlIoWTz1kuTwETzWycmdUCFwFLevVZAlyaaF8I/L7g4+chGDMmuK6nujqYrqqCIUOCdve8sWODsenuMegFCzJfC1RdHVzB3/3cYquq6lnjggXB17Fjg39CetcuIpUt5x66u3eY2TXAI0A1cJe7rzazm4Bl7r4EuBP4iZm1AK8RhH4kLFgQPIr9nMPV6yzIjC6+WAEuElV5jaG7+1Jgaa95N6a09wH/VNjSRESkL/RZLiIiEaFAFxGJCAW6iEhEKNBFRCIi55WiRXths1eB1lBePH+jgL/n7FUeKqlWqKx6VWtxqNb+GevuR6dbEFqgVwIzW5bpEttyU0m1QmXVq1qLQ7UWnoZcREQiQoEuIhIRCvTsFoZdQB9UUq1QWfWq1uJQrQWmMXQRkYjQHrqISEQo0EVEIkKBnsLMvmNmz5vZSjO7z8xGZOi30cxWmdkzZrasxDVWxA27zWy0mTWZ2RozW21m16bpM83MdiW24zNmdmO6dZVKrvfVAv8vsW1XmtlJIdV5fMo2e8bM3jCzz/fqE9q2NbO7zGx74k5m3fNGmtmjZvZC4uuRGZ57aaLPC2Z2abo+Jai17HMgI3fXI/EAzgZqEu1bgFsy9NsIjAqhvmrgRWA8UAs8C0zq1edq4EeJ9kXAz0PalscAJyXaw4D1aWqdBjwY9vue7/sKnAs8DBhwCvBkGdRcDfwvwcUmZbFtgTOBk4DnUuZ9G7g+0b4+3e8WMBLYkPh6ZKJ9ZAi1lnUOZHtoDz2Fu//W3TsSk38luDtTOTl4w253PwB037A71Wzg7kT7l8AMs+577JWOu29z9xWJ9pvAWuDYUtdRYLOBezzwV2CEmR0Tck0zgBfdvWyuunb3xwnui5Aq9efybuAjaZ76YeBRd3/N3XcCjwKzilYo6WutgBzISIGe2eUEe2PpOPBbM1ueuPF1qRwLbE6Z3sKhIXmwT+KHchdwVEmqyyAx7PM+4Mk0i081s2fN7GEze1dJCztUrvc1n+1fahcBP8uwrJy2bZ27b0u0/xeoS9OnHLdvOeZARiW9SXQ5MLPfAW9Ls+gGd38g0ecGoANYlGE1H3D3rWb2VuBRM3s+8ZdeejGzocCvgM+7+xu9Fq8gGCpoM7NzgfuBiaWuMUVFva+JW0JeAHw1zeJy27YHububWdmfL12JORC7PXR3n+nuk9M8usN8LvAPwMWeGChLs46tia/bgfsIhkJKoS837KaoN+zOg5kNIAjzRe7+697L3f0Nd29LtJcCA8xsVInLTK0n1/uaz/YvpXOAFe7+Su8F5bZtgVe6h6cSX7en6VM227fMcyCj2AV6NmY2C/gycIG778nQZ4iZDetuExxAeS5d3yKomBt2J8bt7wTWuvutGfq8rXt838xOJvh5DOuPTz7v6xLgU4mzXU4BdqUMI4RhDhmGW8pp2yak/lxeCjyQps8jwNlmdmTiLJizE/NKqgJyILOwj8qW0wNoIRjDeybx6D5b5O3A0kR7PMHZJc8CqwmGakpZ47kEZ4y82P3awE0EP3wARwC/SHwvfwPGh7QtP0AwxrgyZXueC1wJXJnoc01iGz5LcPDptBDf+7Tva696Dbgtse1XAVNCrHcIQUAPT5lXFtuW4I/MNqCdYBz8CoLjOI8BLwC/A0Ym+k4B7kh57uWJn90W4LKQai37HMj00KX/IiIRoSEXEZGIUKCLiESEAl1EJCIU6CIiEaFAFxGJCAW6iEhEKNBFRCLi/wD8fjLLj0ea6gAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8bG8f8KvOCYm"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}