NYU_Fish_Market_Mini_Project.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "NYU Fish Market Mini Project",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.7.0"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "EjOoKJazwe0W",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.linear_model import LinearRegression\n",
        "from sklearn.linear_model import Ridge\n",
        "from sklearn.linear_model import Lasso\n",
        "from sklearn import model_selection\n",
        "from sklearn import preprocessing"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "KR-q6CIawe0a",
        "colab": {}
      },
      "source": [
        "feature = pd.read_csv('https://raw.githubusercontent.com/asarmadi/summer_ml_s3/master/day05/fish_market_feature.csv')\n",
        "label = pd.read_csv('https://raw.githubusercontent.com/asarmadi/summer_ml_s3/master/day05/fish_market_label.csv')\n",
        "X = feature.values\n",
        "y = label.values"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Lgm7yGyjPXYy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def design_matrix(x, M):\n",
        "    # x = x.reshape(-1,5,1)\n",
        "    # PhiX = np.power(x, np.arange(M+1))\n",
        "    # return PhiX\n",
        "    xshape = x.shape\n",
        "    x_new = x\n",
        "    for i in range(2, M+1):\n",
        "        temp = np.power(x, np.full(shape=xshape, fill_value=i))\n",
        "        x_new = np.hstack([x_new, temp])\n",
        "        # print(i)\n",
        "        # print(x_new.shape)\n",
        "    \n",
        "    PhiX = np.hstack([np.ones(shape=(xshape[0],1)), x_new])\n",
        "    return PhiX"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iIVvJ5Hh99EN",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def RidgeRegression(a, xtrain, xtest, ytrain, ytest):\n",
        "    regr = Ridge(alpha=a, fit_intercept=True).fit(xtrain, ytrain)\n",
        "    yhat_train = regr.predict(xtrain)\n",
        "    yhat_test = regr.predict(xtest)\n",
        "\n",
        "    #print(regr.coef_.shape)\n",
        "\n",
        "    return yhat_train, yhat_test"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kGconmJD_pjE",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# returns array of alpha values that return the lowest MSE value for the testing data set\n",
        "def findBestAlpha(x_train, x_test, y_train, y_test, astart = 0, astop = 1, astep = 0.01):\n",
        "    alpha_values = np.arange(astart, astop, astep)\n",
        "    maetest_values = []\n",
        "    maetrain_values = []\n",
        "\n",
        "    for i in range(0,len(alpha_values)):\n",
        "        yhat_train, yhat_test = RidgeRegression(alpha_values[i], x_train, x_test, y_train, y_test)\n",
        "\n",
        "        MAE_test = np.mean(np.abs(yhat_test - y_test))\n",
        "        MAE_train = np.mean(np.abs(yhat_train - y_train))\n",
        "\n",
        "        maetest_values.append(MAE_test)\n",
        "        maetrain_values.append(MAE_train)\n",
        "\n",
        "        #print(\"Training MAE\", MAE_train)\n",
        "        #print(\"Testing MAE\", MAE_val)\n",
        "\n",
        "    min_train_mae = np.amin(maetrain_values)\n",
        "    min_test_mae = np.amin(maetest_values)\n",
        "\n",
        "    best_train_alpha = alpha_values[np.where(maetrain_values == min_train_mae)]\n",
        "    best_test_alpha = alpha_values[np.where(maetest_values == min_test_mae)]\n",
        "\n",
        "    # print(f\"The minimum MAE training value was {min_train_mae} with an alpha of {best_train_alpha}\")\n",
        "    # print(f\"The minimum MAE testing value was {min_test_mae} with an alpha of {best_test_alpha}\")\n",
        "\n",
        "    # print(\"Y | Y-hat:\")\n",
        "    # Y = np.hstack([y_test.reshape((-1, 1)), yhat_test.reshape((-1, 1))])\n",
        "    # with np.printoptions(precision=2):\n",
        "    #     print(Y[:10,:])\n",
        "\n",
        "    return best_test_alpha"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pi93KZuAhyGj",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 101
        },
        "outputId": "b06d1b79-a4c4-496a-fecb-7344c6c2a311"
      },
      "source": [
        "# normalizing data\n",
        "Xs = preprocessing.scale(X)\n",
        "#y = preprocessing.scale(y)\n",
        "# splitting data\n",
        "X_train, X_val, y_train, y_val = model_selection.train_test_split(Xs, y, test_size=0.5, train_size=0.5, random_state=123)\n",
        "print(X_train.shape)\n",
        "print(y_train.shape)\n",
        "\n",
        "M = 5 #appears to be the best M-value\n",
        "X_train = design_matrix(X_train, M)\n",
        "print(X_train.shape)\n",
        "X_val = design_matrix(X_val, M)\n",
        "# print(X_val.shape)\n",
        "\n",
        "# get the best alpha value\n",
        "validation_alpha = findBestAlpha(X_train, X_val, y_train, y_val, astop=10)[0]\n",
        "print(f\"Best Alpha: {validation_alpha}\")\n",
        "\n",
        "'''\n",
        "I created a design matrix that has M*5+1 columns (from the M order polynomial and 5 features in the data) and 62 rows (from the original data)\n",
        "The design matrix looks like (for a dataset with only 2 features x1,x2 and M=3):\n",
        "[ 1 + x1 + x1^2 + x1^3 + x2 + x2^2 + x2^3 ]\n",
        "[ ... ...  ...    ...    ...  ...    ...  ]\n",
        "\n",
        "\n",
        "The weights for each value are found in the sklearn Ridge regression function\n",
        "Used to calculate the 26 different weights (for a M=5) which made the alpha value non-zero\n",
        "The function for outputting yhat looks like (for a dataset with only 2 features x1,x2 and M=3): \n",
        "yhat = w0 + w1*x1 + w2*x1^2 + w3*x1^3 + w4*x2 + w5*x2^2 + w6*x2^3 ]\n",
        "\n",
        "*However since the alpha is non-zero some of the weights that are removed due to the regularization (I think)\n",
        "\n",
        "'''\n",
        "print()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(62, 5)\n",
            "(62, 1)\n",
            "(62, 26)\n",
            "Best Alpha: 0.04\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5MpamsTQ6eH6",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 50
        },
        "outputId": "2a4c2403-f7cf-4034-8c08-ad4e651a86bc"
      },
      "source": [
        "''' \n",
        "The test set will be released on Sunday night\n",
        "'''\n",
        "X_test = pd.read_csv('https://raw.githubusercontent.com/asarmadi/summer_ml_s3/master/day05/fish_market_test_feature.csv').values\n",
        "y_test = pd.read_csv('https://raw.githubusercontent.com/asarmadi/summer_ml_s3/master/day05/fish_market_test_label.csv').values\n",
        "# print(X_test[:,0])\n",
        "X = preprocessing.scale(X_test)\n",
        "# print(X_test[:,0])\n",
        "#y_test = preprocessing.scale(y_test)\n",
        "\n",
        "X = design_matrix(X, M)\n",
        "yhat_train, yhat_test = RidgeRegression(validation_alpha, X_train, X, y_train, y_test)\n",
        "\n",
        "mae_test = np.mean(np.abs(yhat_test - y_test))\n",
        "mae_train = np.mean(np.abs(yhat_train - y_train))\n",
        "\n",
        "print(f\"Training MAE: {mae_train}\")\n",
        "print(f\"Testing MAE: {mae_test}\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training MAE: 14.329912981486665\n",
            "Testing MAE: 44.34968861804575\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nMIzdJsd1Qjh",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "882ea319-2438-4410-8f2b-58ff6f1028aa"
      },
      "source": [
        "for i in range(0,X_test.shape[1]):\n",
        "    plt.figure()\n",
        "    plt.scatter(X_test[:,i], y_test)\n",
        "    plt.scatter(X_test[:,i], yhat_test)\n",
        "    plt.legend([\"Actual Values\", \"Predicted Values\"])\n",
        "    plt.xlabel(f\"{feature.columns.values[i]}\")\n",
        "    plt.ylabel(f\"{label.columns.values[0]}\")\n",
        "    plt.title(f\"Feature: {i+1}\")\n",
        "\n",
        "plt.figure()\n",
        "plt.scatter(y_test, yhat_test)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.collections.PathCollection at 0x7fd9f1e7ee10>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 17
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEWCAYAAACXGLsWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3de3xU1b338c+PAE28EYUUhUCDPR4qhhgwgAhaFC1egWJFeU6rtiDtY6loj1g8r1apx1aUp1Kh51Sp9qitBRERr60XLo/WViXIRUBBnoqSiIogVCRISH7PH7MHJ8kkM0lmMpd836/XvGb22nv2/mWj+WWttdda5u6IiIg0pUOqAxARkfSnZCEiIjEpWYiISExKFiIiEpOShYiIxKRkISIiMSlZiIhITEoW0m6Z2VYzqzKzvRGvHgk459mJirEF1y82s2fN7GMz0yAqSRglC2nvLnL3IyJe76cyGDPr2MpTVAMLgYkJCEfkECULkXrMrIuZ3Wdm282s0sxuNbOcYN9XzWyZme0M/np/yMzyg31/AHoDTwa1lBvMbISZVdQ7/6Hah5nNMLNFZvZHM/sncGVT14/F3Te5+33AhkTeExElC5GG7gcOAv8CDAC+AUwK9hlwG9ADOBHoBcwAcPfvAO/xRW3ljjivNwZYBOQDDzV1fTPrbWa7zax3a35AkeZSspD2bknwy3e3mS0xs+7A+cC17v6Zu38EzAYuA3D3Le7+vLt/7u47gDuBr7cyhr+7+xJ3rwWOinH999w9393fa+U1RZqlte2jIplurLu/EN4ws8FAJ2C7mYWLOwDbgv3dgbuA04Ejg32ftDKGbRGfv9LU9UVSRclCpK5twOdAN3c/GGX/LwEH+rv7LjMbC/wmYn/9J5A+Aw4LbwR9DwX1jon8Tqzri6SEmqFEIrj7duA54FdmdpSZdQg6tcNNTUcCe4E9ZtYTmFbvFB8Cx0dsbwZyzewCM+sE/BT4Uiuu3yQLyQU6B9u5Ztbo9UTipWQh0tDlhH7ZbiTUxLQIOC7Y93NgILAHeBpYXO+7twE/DfpArnf3PcDVwL1AJaGaRgVNa/T6QQf33iY6uL8CVPHF01BVwKZYP7BILKbFj0REJBbVLEREJCYlCxERiUnJQkREYlKyEBGRmLJynEW3bt28qKgo1WGIiGSUVatWfezu9ccBAVmaLIqKiigvL091GCIiGcXM3m1sn5qhREQkJiULERGJSclCRERiyso+i2iqq6upqKhg//79qQ5FEig3N5fCwkI6deqU6lBEslq7SRYVFRUceeSRFBUVETH1s2Qwd2fnzp1UVFTQp0+fVIcjktXaTTPU/v376dq1qxJFFjEzunbtqtpiNlu3EGYXw4z80Pu6hamOqN1qNzULQIkiC+nfNIutWwhPXgPVVaHtPdtC2wAl41MXVzvVbmoWIpJhlt7yRaIIq64KlUubU7JoY0uWLMHMeOutt2Ie++tf/5p9+/a1+Fr3338/U6ZMqVO2detWCgsLqa2trVNeWlrKq6++GvU8W7dupbi4uMVxiLTInkaW/WisvAWWrK5k2Mxl9Jn+NMNmLmPJ6sqEnTvbKFm0sfnz5zN8+HDmz58f89jWJotoioqK6N27Ny+99NKhsrfeeotPP/2UIUOGJPRaIq3SpbB55c20ZHUlNy5+g8rdVThQubuKGxe/oYTRCCWLRiTjL469e/fy17/+lfvuu48FCxYcKq+pqeH666+nuLiYkpIS5s6dy5w5c3j//fc588wzOfPMMwE44ogjDn1n0aJFXHnllQA8+eSTDBkyhAEDBnD22Wfz4YcfNhnHhAkT6lx/wYIFXHbZZWzdupXTTz+dgQMHMnDgQP72t781+G792sqFF17IihUrAHjuuecYOnQoAwcO5JJLLmHv3r0ATJ8+nX79+lFSUsL111/fvJsm7dfIm6BTXt2yTnmh8gSY9ewmqqpr6pRVVdcw61ktLBhNu+rgjlf4L47wf0jhvzgAxg7o2eLzPv7445x77rn867/+K127dmXVqlWccsopzJs3j61bt7JmzRo6duzIrl27OOaYY7jzzjtZvnw53bp1a/K8w4cP55VXXsHMuPfee7njjjv41a9+1ejx48ePp7S0lLlz59KxY0cefvhhHnnkEb785S/z/PPPk5uby9tvv82ECRPinmPr448/5tZbb+WFF17g8MMP5/bbb+fOO+/khz/8IY899hhvvfUWZsbu3bubdc+kHQt3Yi+9JdT01KUwlCgS1Ln9/u6qZpW3d0oWUTT1F0drksX8+fOZOnUqAJdddhnz58/nlFNO4YUXXuAHP/gBHTuG/jmOOeaYZp23oqKCSy+9lO3bt3PgwIGYYw66d+9OcXExS5cupXv37nTs2JHi4mL27NnDlClTWLNmDTk5OWzevDnuGF555RU2btzIsGHDADhw4ABDhw6lS5cu5ObmMnHiRC688EIuvPDCZv1s0s6VjE/ak0898vOojJIYeuTnRTlalCyiSMZfHLt27WLZsmW88cYbmBk1NTWYGbNmzYr7HJGPiUaOLfjRj37Ej3/8Y0aPHs2KFSuYMWNGzHOFm6K6d+/OhAkTAJg9ezbdu3dn7dq11NbWkpub2+B7HTt2rNM5Ho7D3TnnnHOi9sW89tprLF26lEWLFvGb3/yGZcuWxf0ziyTLtFF967QgAOR1ymHaqL4pjCp9qc8iisb+smjNXxyLFi3iO9/5Du+++y5bt25l27Zt9OnTh5deeolzzjmHe+65h4MHDwKhxAJw5JFH8umnnx46R/fu3XnzzTepra3lscceO1S+Z88eevYM1XgeeOCBuOIZN24czzzzDA8//DCXXXbZofMcd9xxdOjQgT/84Q/U1NQ0+F5RURFr1qyhtraWbdu28dprrwFw6qmn8vLLL7NlyxYAPvvsMzZv3szevXvZs2cP559/PrNnz2bt2rXNvXUiSTF2QE9uG9efnvl5GNAzP4/bxvVvVetBNlPNIopk/MUxf/58fvKTn9Qpu/jii5k/fz5z585l8+bNlJSU0KlTJ6666iqmTJnC5MmTOffcc+nRowfLly9n5syZXHjhhRQUFFBWVnaoA3nGjBlccsklHH300Zx11lm88847MePJz89n6NChfPDBBxx//PEAXH311Vx88cU8+OCDnHvuuRx++OENvjds2DD69OlDv379OPHEExk4cCAABQUF3H///UyYMIHPP/8cgFtvvZUjjzySMWPGsH//ftydO++8s8X3UCTRxg7oqeQQJ3P3VMeQcGVlZV6/Y/bNN9/kxBNPjPscS1ZXMuvZTby/u4oe+XlMG9VX/1Glqeb+24pIdGa2yt3Lou1TzaIR+otDROQL6rMQEZGYlCxERCQmJQsREYlJyUJEJBqtpVGHOrhFROrTWhoNqGbRhnJycigtLaW4uJhLLrmkVTPKXnnllSxatAiASZMmsXHjxkaPXbFiRdRJAWMpKiri448/rlP23e9+l3vuuadO2ZIlSzjvvPPiilUkI2gtjQaULNpQXl4ea9asYf369XTu3Jm77767zv7wCO7muvfee+nXr1+j+1uaLKKpP2MthGatDU8ZIpIV2mAtjUyTtGRhZr83s4/MbH1E2TFm9ryZvR28Hx2Um5nNMbMtZrbOzAZGfOeK4Pi3zeyKZMXbQJLbK08//XS2bNnCihUrOP300xk9ejT9+vWjpqaGadOmMWjQIEpKSg79Fe/uTJkyhb59+3L22Wfz0UcfHTrXiBEjDs0O+5e//IWBAwdy8sknM3LkSLZu3crdd9/N7NmzKS0t5aWXXmLHjh1cfPHFDBo0iEGDBvHyyy8DsHPnTr7xjW9w0kknMWnSJKIN2Bw5ciRvvfUW27dvB0LTerzwwguMHTuWW265hUGDBlFcXMzkyZOjfj+ytlJeXs6IESMOned73/segwcPZsCAATz++OMAbNiwgcGDB1NaWkpJSQlvv/12gv4FRJqQ5LU0MlEyaxb3A+fWK5sOLHX3E4ClwTbAecAJwWsy8FsIJRfgZmAIMBi4OZxgkircXrlnG+BftFcmKGEcPHiQP//5z/Tv3x+A119/nbvuuovNmzdz33330aVLF1auXMnKlSv53e9+xzvvvMNjjz3Gpk2b2LhxIw8++GDUmsKOHTu46qqrePTRR1m7di2PPPIIRUVF/OAHP+C6665jzZo1nH766UydOpXrrruOlStX8uijjzJp0iQAfv7znzN8+HA2bNjAN7/5Td57770G18jJyeHiiy9m4cLQvXjyyScZMWIERx11FFOmTGHlypWsX7+eqqoqnnrqqbjvyS9+8QvOOussXnvtNZYvX860adP47LPPuPvuu5k6dSpr1qyhvLycwsL2+z+rtJ2VX/0RVd65TlmVd2blV3+UoohSL2nJwt1fBHbVKx4DhGe6ewAYG1H+oIe8AuSb2XHAKOB5d9/l7p8Az9MwASVektorq6qqKC0tpaysjN69ezNx4kQABg8efGha8eeee44HH3yQ0tJShgwZws6dO3n77bd58cUXmTBhAjk5OfTo0YOzzjqrwflfeeUVzjjjjEPnamyq8xdeeIEpU6ZQWlrK6NGj+ec//8nevXt58cUX+fa3vw3ABRdcwNFHR8/LkU1RkU1Qy5cvZ8iQIfTv359ly5axYcOGuO/Nc889x8yZMyktLWXEiBHs37+f9957j6FDh/LLX/6S22+/nXfffZe8PE0fLcl37cYT+En1JCpqu1HrRkVtN35SPYlrN56Q6tBSpq2fhuru7tuDzx8A3YPPPYFtEcdVBGWNlTdgZpMJ1Uro3bt366JMUntluM+ivsgJ+9yduXPnMmrUqDrHPPPMM626dqTa2lpeeeWVqFOQx+O0005j+/btrF27lr/97W8sWLCA/fv3c/XVV1NeXk6vXr2YMWNGnWnUwyKnOI/c7+48+uij9O1bd7LGE088kSFDhvD0009z/vnnc88990RNlCKJ9P7uKioZzhMHhtcpt3a8MFLKOrg91KCdsFkM3X2eu5e5e1lBQUHrTpbC9spRo0bx29/+lurqagA2b97MZ599xhlnnMHDDz9MTU0N27dvZ/ny5Q2+e+qpp/Liiy8emnW2sanOv/GNbzB37txD2+EEdsYZZ/CnP/0JgD//+c988sknUWM0My699FKuuOIKzjvvPHJzcw/94u/WrRt79+5t9OmnoqIiVq1aBcCjjz5a5+eeO3fuoX6O1atXA/CPf/yD448/nmuuuYYxY8awbt26Ju+fSLM00jeZjGUKMl1bJ4sPg+YlgvdwL20l0CviuMKgrLHy5Ery2r9NmTRpEv369WPgwIEUFxfz/e9/n4MHD/LNb36TE044gX79+nH55ZczdOjQBt8tKChg3rx5jBs3jpNPPplLL70UgIsuuojHHnvsUAf3nDlzKC8vp6SkhH79+h16Kuvmm2/mxRdf5KSTTmLx4sVN1tAmTJjA2rVrDzVB5efnc9VVV1FcXMyoUaMYNGhQ1O/dfPPNTJ06lbKyMnJycg6V/+xnP6O6upqSkhJOOukkfvaznwGwcOFCiouLKS0tZf369Vx++eUtu7Ei9TXRNzltVF/yOuXUOby9L4yU1CnKzawIeMrdi4PtWcBOd59pZtOBY9z9BjO7AJgCnE+oM3uOuw8OOrhXAeGno14HTnH3+n0hdSRiinLWLUza2r+SWJqiXFpkdnGQKOrp0guuW98ulylIyRTlZjYfGAF0M7MKQk81zQQWmtlE4F0g/Nv3GUKJYguwD/gugLvvMrP/BFYGx90SK1EkTBLX/hWRNBCjb1LLFNSVtGTh7o2N0hoZ5VgHftjIeX4P/D6BoYmIhFoMotYs9Hh2NO1qBHc2rgrY3unfVFoshX2TmajdJIvc3Fx27typXy5ZxN3ZuXNnix8BlhRJl9lcS8bDRXNCfRRY6P2iOWp+bkS7mXW2sLCQiooKduzYkepQJIFyc3M1qjuTpNtsruqbjFu7SRadOnU6NLJZRFKkqdkR9Es7rbWbZigRSQOazTVjtZuahYikgWY+gdQexzqkK9UsRKTtjLyJgzl1H0g4mJMb9QmkJasruXHxG1TursKByt1V3Lj4DZasTv4kDtKQkoWItJklNcOYXm821+nVk1hSM6zBsbOe3URVdU2dsqrqGmY9u6mtwpUIaoYSkTYz69lNVB44jUWcVqf8789uatC89H4jM7w2Vi7JpZqFiLSZ5iQAzfyaXpQsRKTNNCcBaObX9KJkISJtpjkJYOyAntw2rj898/MwoGd+HreN66+noVJEfRYi0mbCv+jjfRxWM7+mDyULEWlTSgCZSc1QIiISk5KFiIjEpGQhIiIxKVmIiEhMShYiIhKTkoWIiMSkZCEiIjEpWYiISExKFiIiEpOShYiIxKRkISIiMSlZiIhITEoWIiISU0qShZldZ2YbzGy9mc03s1wz62Nmr5rZFjN72Mw6B8d+KdjeEuwvSkXMIiLtWZsnCzPrCVwDlLl7MZADXAbcDsx2938BPgEmBl+ZCHwSlM8OjhMRkTaUqmaojkCemXUEDgO2A2cBi4L9DwBjg89jgm2C/SPNzNowVhGRdq/Nk4W7VwL/B3iPUJLYA6wCdrv7weCwCiC8OkpPYFvw3YPB8V3rn9fMJptZuZmV79ixI7k/hIhIO5OKZqijCdUW+gA9gMOBc1t7Xnef5+5l7l5WUFDQ2tOJiEiEVDRDnQ284+473L0aWAwMA/KDZimAQqAy+FwJ9AII9ncBdrZtyCIi7Vsq1uB+DzjVzA4DqoCRQDmwHPgWsAC4Ang8OP6JYPvvwf5l7u5tHbSISDpbsrqSWc9u4v3dVfTIz2PaqL4JXeu8zZOFu79qZouA14GDwGpgHvA0sMDMbg3K7gu+ch/wBzPbAuwi9OSUiIgElqyu5MbFb1BVXQNA5e4qblz8BkDCEoZl4x/pZWVlXl5enuowRETaxLCZy6jcXdWgvGd+Hi9PPyvu85jZKncvi7ZPI7hFRDLc+1ESRVPlLaFkISKS4Xrk5zWrvCWULEREMty0UX3J65RTpyyvUw7TRvVN2DVS8TSUiIgkULgTO6uehhIRkcQbO6BnQpNDfWqGEpGQdQthdjHMyA+9r1uY6ogkjahmISKhxPDkNVAdPD2zZ1toG6BkfOrikrShmoWIwNJbvkgUYdVVoXIRlCxEBGBPRfPKpd1RM5RIlmnRHEFdCkNNT9HKRVDNQiSrhOcIqtxdhfPFHEFLVlc2/cWRN0GnegO4OuWFykVQshDJKrOe3XRoMrmwquoaZj27qekvloyHi+ZAl16Ahd4vmqPObTlEzVAiWaRVcwSVjFdykEapZiGSRdpijiBJU0keJ6NkIZJF2mKOIElD4XEye7YB/sU4mQQmDCULkSwydkBPbhvXn575eRih9QxuG9c/qdNASBpog3Ey6rMQyTLJniNI0lAbjJNRzUIkk2j+JommsfEwCRwno2QhkikS1S6thJN92mCcjJKFSKZIRLt0G3SESgq0wTgZ9VmIZIpEtEs3lXA0xiKzJXmcjGoWIpkiEe3SmjBQWiiuZGFmt8dTJiJJlIh26TboCJXsFG/N4pwoZeclMhARiSER7dKaMFBaqMk+CzP738DVwPFmti5i15HAy8kMTESiaG27dPi7S28JNT11KQwlCvVXSAyxOrj/BPwZuA2YHlH+qbvvSlpUIpI8mjBQWqDJZOHue4A9wAQzywG6B985wsyOcPf32iBGERFJsbgenTWzKcAM4EOgNih2oKQlFzWzfOBeoDg4z/eATcDDQBGwFRjv7p+YmQF3AecD+4Ar3f31llxXJNO1aBU8kQSIt4P7WqCvu5/k7v2DV4sSReAu4C/u/jXgZOBNQs1cS939BGApXzR7nQecELwmA79txXVFMlaLV8ETSYB4k8U2Qs1RrWZmXYAzgPsA3P2Au+8GxgAPBIc9AIwNPo8BHvSQV4B8MzsuEbGIZJK4V8HTdB6SBLGehvpx8PEfwAozexr4PLzf3e9swTX7ADuA/zGzk4FVwFSgu7tvD475gFD/CEBPQskqrCIo2x5RhplNJlTzoHfv3i0ISyS9xbUKXng6j/Ao7fB0HqBObWmVWDWLI4PXe8DzQOeIsiNbeM2OwEDgt+4+APiMuk9a4e5OqC8jbu4+z93L3L2soKCghaGJpK+4VsGLY/6oJasrGTZzGX2mP82wmcvUjCVxifU01M+TcM0KoMLdXw22FxFKFh+a2XHuvj1oZvoo2F8J9Ir4fmFQJtKuTBvVlxsXv1GnKarBKngxpvMI93uEzxHu9wDUUS5Nine6jyfN7Il6rz+Y2VQzy23OBd39A2CbmYX/Cx8JbASeAK4Iyq4AHg8+PwFcbiGnAnsimqtE2o24VsGLMZ1H3P0eIvXEO+vsP4ACYH6wfSnwKfCvwO+A7zTzuj8CHjKzzsG5v0socS00s4nAu0C4gfUZQo/NbiH06Ox3m3ktkawRcxW8kTfV7bOAOtN5xNXvIRJFvMniNHcfFLH9pJmtdPdBZrahuRd19zVAWZRdI6Mc68APm3sNkXYpxnQePfLzqIySGHrk52kMhzQp3mRxhJn1Do/YNrPewBHBvgNJiUxEWqaJ6Twa6/c482sF6suQJsU7zuLfgb+a2XIzWwG8BFxvZofzxdgIEYG0HufQWL/H8rd2qC9DmhRXzcLdnzGzE4CvBUWb3H1/8PnXSYlMJBNlwDiHaP0e1z28Juqx6suQsCZrFmZ2VvA+DrgA+GrwOj8oE5FIiVgnOwXiGsMh7VqsZqivB+8XRXldmMS4RDJThi5bOm1UX/I65dQpazCGQ9q1WIPybg7e9biqSDy6FIaanqKVp7Fws5SehpLGxDtFeXfgl0APdz/PzPoBQ939vqRGJ5JpYoxzSGcxx3BIuxbv01D3A88CPYLtzYSmLReRSIlYJ1skDcU7zqKbuy80sxsB3P2gmdXE+pJIu6RlSyULxZssPjOzrgQzwYbnaEpaVCIZTCOhJRvFWs/iWuBvwA2EJvY73sxeJjRP1CXJD08ks2hWV8lWsfosCgkNuvtLcOzzwAJCc0WtTXJsIhlHs7pKtor16Oz1AMHssGXAacAI4EYz2+3u/ZIeoUgG0ayukq3ifRoqDzgK6BK83gdebfIbIu2QRkJLtoo13ce8oI/iYWAoof6LS4LlSzVQT6SeNhkJncYTFUr2ivU0VG/gS8DbhJYyrQB2JzsokUyV9JHQGTBRoWQnC60t1MQBZgacRKi/4jSgGNgF/D08HUi6KSsr8/Ly8lSHIZJ4s4sbmU6kF1y3vu3jkaxiZqvcPdrCdLHHWQQr1a03s92ExlbsITSJ4GAgLZOFSNbK0IkKJfPF6rO4xswWmNl7wP8llCTeAsYBx7RBfCISqbEJCdN8okLJfLFqFkXAI8B17r49+eGISJMyeKJCyWyxxln8uK0CEZE4hDuxl94SanrqUhhKFOrcliSLd24oEUkXmqhQUiDeQXkiItKOKVmIiEhMShYiIhKTkoWIiMSkZCEiIjGlLFmYWY6ZrTazp4LtPmb2qpltMbOHg2nRMbMvBdtbgv1FqYpZRKS9SmXNYirwZsT27cBsd/8X4BNgYlA+EfgkKJ8dHCciIm0oJcnCzAqBC4B7g20DzgIWBYc8AIwNPo8Jtgn2jwyOF2mXlqyuZNjMZfSZ/jTDZi5jyerKVIck7UCqaha/JrSud22w3RXY7e4Hg+0KIDync09gG0Cwf09wfB1mNtnMys2sfMeOHcmMXSRlwmt8V+6uwvlijW8lDEm2Nk8WZnYh8JG7r0rked19XrAoU1lBQUEiTy2SNrTGt6RKKqb7GAaMNrPzgVxCy7XeBeSbWceg9lBIaLElgvdeQIWZdSS0rOvOtg9bJPW0xrekSpvXLNz9RncvdPci4DJgmbv/G7Ac+FZw2BXA48HnJ4Jtgv3LPNaKTSJZSmt8S6qk0ziLnwA/NrMthPok7gvK7wO6BuU/BqanKD6RlGuTNb5FokjprLPuvgJYEXz+B6HV9+ofsx+4pE0DE0lTSV/jW6QRmqJcJMOMHdBTyUHaXDo1Q4mISJpSshARkZiULETitW4hzC6GGfmh93ULUx2RSJtRn4VIPNYthCevgepgPMOebaFt0BKn0i6oZiESj6W3fJEowqqrQuUi7YCShUg89lQ0r1wkyyhZSHZJVr9Cl8LmlYtkGSULyR7hfoU92wD/ol8hEQlj5E0czMmtU3QwJxdG3tT6c4tkACULyR5J7FdYUjOM6dWTqKjtRq0bFbXdmF49iSU1w1p9bpFMoKehJHvE2a+w8ol76PX6LL7sO/jICtg2cBqDRn+/yVPPenYTlQdOYxGnATC6w1+5oeMCejz+X7CiMFTD0FNRksVUs5DsEUe/wson7qF41U85lh10MDiWHRSv+ikrn7inyVNHTgE+usNfmdnpXgo7fEyHRDd3iaQpJQvJHiNvgk71purulFenX6HX67PIswN1DsmzA/R6fVaTp46cAvyGjgs5rN459BitZDslC8keJeNZ2f/nfEABtW58QAEr+/+8TvPQlz36krtf9o+bPHXk1OA9rJFj9RitZDH1WUjWWLK6khtXfoWq6rsOleWtzOG2XpWHZmn9yAo4loYJ4yPrxrFNnLvO1OD7ulEYLWHoMVrJYqpZSNaIZ33qbQOnUeWd6x7jndk2cFrM848d0JOXp59F4bdui9ncJZJtlCwka8SzPvWg0d9n/Sm31mmqWn/KrTGfhqqjZDxcNAe69AIs9H7RHD0NJVlNzVCSNXrk51EZJWHUX5960OjvQ5Acjg1ezVYyXslB2hXVLCRrtOX61EtWVzJs5jL6TH+aYTOXsWR1ZcKvIZJOVLOQrNFW61MvWV3JjYvfONQ/Urm7ihsXv1EnBpFso2QhWSXq+tTrFobGQOypCD2x1MrR1k11pCtZSLZSspDsloRFi+LpSBfJNuqzkOyWhMkF63eYxyoXyQZKFpLdkrBoUVt2pIukCzVDSXbrUhisbxGlvIXaqiNdJJ0oWUh2G3lT3T4LSMho66gd6SJZrM2bocysl5ktN7ONZrbBzKYG5ceY2fNm9nbwfnRQbmY2x8y2mNk6MxvY1jFLBtNoa5GESEWfxUHg3929H3Aq8EMz6wdMB5a6+wnA0q4z3yQAAAxjSURBVGAb4DzghOA1Gfht24csmWxJzTCGfT6HPvsfYtjnc7S6nUgLtHmycPft7v568PlT4E2gJzAGeCA47AFgbPB5DPCgh7wC5JvZcW0ctmSo8AC6yt1VOF8MoNOIa5HmSenTUGZWBAwAXgW6u/v2YNcHQPfgc08gsoeyIigTiSmemWhFJLaUJQszOwJ4FLjW3f8Zuc/dHfBmnm+ymZWbWfmOHdEXuJH2RwPoRBIjJcnCzDoRShQPufvioPjDcPNS8P5RUF4J9Ir4emFQVoe7z3P3MncvKygoSF7w0nbWLYTZxTAjP/TegjWuNYBOJDFS8TSUAfcBb7r7nRG7ngCuCD5fATweUX558FTUqcCeiOYqyVSxEsG6hRx8/EfBGAmHPdtC281MGBpAJ5IYqRhnMQz4DvCGma0Jyv4DmAksNLOJwLtA+NnGZ4DzgS3APuC7bRuuJETkZH55R8OBvVBzILQvynxN+/58E4fV7K9zio41+0PlzXjsVQPoRBKjzZOFu/8VsEZ2j4xyvAM/TGpQklz1J/Or2tXwmOoqeOwHoc8l48mt+iDqqRorb4oG0Im0nuaGkuSLNplfNF5D7eKr8Ju7UOvR/554v7ZrgoMTkXgoWUjyNWPSvg6AGXS0Wrze83D7vDP3dv52YmMTkbgoWUjytXDSPjM46B2odaOiths3+WRKL5ic4OBEJB5KFpJ8I28KTd7XAh1wvvr5Q1x62O8Y/s2r1fcgkiKadVaSL/z0UvA0lOONPuFQ30fWjXdmXpC00EQkPqpZSGLEGjdRMh6uWw8zdrMvp0vUU9Tvo6jyzmwbOC1JAYtIcyhZSOuFH42NGEDHk9c0OoBuln2XA163UnvAO/JgzdlU1Haj1o0PKGD9KbcyaPT32+AHEJFY1AwlrdfUOtdRBtA9sHcwuzpM5oaOC+lhO3nfu3LHwfE8UTucLpeWMnZAT44Fjm2b6EUkDkoW0npxrHO9ZHXloVHUHcx4onY4TxwYXufwnvl56sAWSVNqhpJW25cXvQ4QLq+/pkRN/c4JNF+TSLpTspCQVszwekf1pezzznXK9nln7qi+FIi+pgRAjhlGqEZx27j+qlWIpDE1Q0nDuZuiTOzXlFAfxIEGfRBPfj6YGTS+dkStux6LFckQShbS7A7q+nrk5/HE7uh9EOH9lVEShtaUEMkcaobKNi1pToqjg7opsdaM0JoSIplPNYts0sLmpH15x3JYVcP1pPblHcthcVw21poRWlNCJPMpWWSTGM1JkY+vRv7CvqP6Um7w/+YwO3Doa+EO6hlxXjrWmhFaU0IksylZZBHfUxF1ziXfU8Hjqyv562P/zcMsoMeXPub9fd349WOXAVfH7KAWEVGyyCIf0o1j2RG1fM3T87jF5h2qPRTax9zi87jj6Y70yP96kx3UIiLq4M4itx24JOp4h9sOXMKkA3+s08wEcJgdYNKBP6oDWkRiUrLIIuVHncP06kmHJuOrqO3G9OpJlB91Dj067Iz6nR4ddjJ2QE9uG9efnvl5GiQnIlGpGSqLTBvVlxsXH6jTnJTXKYfbRvVl/3PRn3jaHzzxpA5oEWmKahZZpKkawmHn3cLBnNw6xx/MyeWw825JTbAiklFUs8gyjdYQSsaH/rGD1eroUkjHkTfFNUJbRETJItK6hXV+mZJtv0xLxmfXzyMibUbJIqwlo5+zPbmIiASULMKaGv0c3h+ZFKBVM7WKiGQS8ygL0WS6srIyLy8vb96XZuQD0e/FwZxcOtbsr7vd+TCo2tXw4C694Lr1zbu2iEgaMLNV7l4WbV/GPA1lZuea2SYz22Jm0xN9/sZWe6uhQ51EAdCxZj8eLVEQmlpDRCTbZESyMLMc4L+A84B+wAQz65fIazS22pt5bfQvNFIh+5BuiQxLRCQtZESyAAYDW9z9H+5+AFgAjEnkBR7YOzjq6Of3Pfov/11+RKNTa4iIZJtM6eDuCWyL2K4AhkQeYGaTgckAvXv3bvYFGlvtrUON8Uv7XYPpu/+z5gpq3RvM1LrqqHOafW0RkXSXKTWLmNx9nruXuXtZQUFBs7/f2GR6RwyawE0+uU6N4yafzBGDJvB8ztcZfmAOx3/+EMMPzOH5nK9r8j0RyUqZUrOoBHpFbBcGZQnT1GpuS75yDJc+O7JBedlXjtHqbyLSLmTEo7Nm1hHYDIwklCRWAv/L3TdEO75Fj86KiLRzTT06mxE1C3c/aGZTgGeBHOD3jSUKERFJvIxIFgDu/gzwTKrjEBFpj7Kmg1tERJJHyUJERGJSshARkZgy4mmo5jKzHcC7rThFN+DjBIWTbJkUKyjeZMqkWEHxJltL4v2Ku0cdqJaVyaK1zKy8scfH0k0mxQqKN5kyKVZQvMmW6HjVDCUiIjEpWYiISExKFtHNS3UAzZBJsYLiTaZMihUUb7IlNF71WYiISEyqWYiISExKFiIiElO7ThZm9nsz+8jM1keUHWNmz5vZ28H70amMMVIj8c4ws0ozWxO8zk9ljGFm1svMlpvZRjPbYGZTg/K0vL9NxJuu9zfXzF4zs7VBvD8PyvuY2avBWvUPm1nnWOdqC03Ee7+ZvRNxf0tTHWuYmeWY2WozeyrYTst7GxYl3oTe23adLID7gXPrlU0Hlrr7CcDSYDtd3E/DeAFmu3tp8EqXyRYPAv/u7v2AU4EfBuump+v9bSxeSM/7+zlwlrufDJQC55rZqcDthOL9F+ATYGIKY4zUWLwA0yLu75rUhdjAVODNiO10vbdh9eOFBN7bdp0s3P1FYFe94jHAA8HnB4CxbRpUExqJNy25+3Z3fz34/Cmh/4h7kqb3t4l405KH7A02OwUvB84CFgXl6XR/G4s3LZlZIXABcG+wbaTpvYWG8SZDu04Wjeju7tuDzx8A3VMZTJymmNm6oJkqLZp1IplZETAAeJUMuL/14oU0vb9Bs8Ma4CPgeeD/Abvd/WBwSAVplPDqx+vu4fv7i+D+zjazL6UwxEi/Bm4AaoPtrqTxvaVhvGEJu7dKFk3w0HPFafvXT+C3wFcJVe23A79KbTh1mdkRwKPAte7+z8h96Xh/o8SbtvfX3WvcvZTQMsODga+lOKQm1Y/XzIqBGwnFPQg4BvhJCkMEwMwuBD5y91WpjiUeTcSb0HurZNHQh2Z2HEDw/lGK42mSu38Y/E9YC/yO0C+NtGBmnQj94n3I3RcHxWl7f6PFm873N8zddwPLgaFAfrAMMSRhrfpEiIj33KD5z939c+B/SI/7OwwYbWZbgQWEmp/uIn3vbYN4zeyPib63ShYNPQFcEXy+Ang8hbHEFP7FG/gmsL6xY9tS0MZ7H/Cmu98ZsSst729j8abx/S0ws/zgcx5wDqF+luXAt4LD0un+Rov3rYg/HIxQH0DK76+73+juhe5eBFwGLHP3fyNN720j8X470fc2Y5ZVTQYzmw+MALqZWQVwMzATWGhmEwlNcz4+dRHW1Ui8I4JH4hzYCnw/ZQHWNQz4DvBG0E4N8B+k7/1tLN4JaXp/jwMeMLMcQn/0LXT3p8xsI7DAzG4FVhNKgOmgsXiXmVkBYMAa4AepDDKGn5Ce97YxDyXy3mq6DxERiUnNUCIiEpOShYiIxKRkISIiMSlZiIhITEoWIiISk5KFSAxmtjf2Ua06/7Vmdlis65nZGWb2upkdNLNvRTtGJFmULERS71rgsJhHwXvAlcCfkhqNSBTtelCeSEuZ2VeB/wIKgH3AVe7+lpndD/wTKAOOBW5w90Vm1gH4DaGpI7YB1cDvgR7Ba7mZfezuZwbn/wVwIVAFjAmmHdka7Ks/WZxI0qlmIdIy84AfufspwPXAf0fsOw4YTuiX/cygbBxQBPQjNFJ8KIC7zwHeB84MJwrgcOCVYO2HF4GrkvqTiMRBNQuRZgpmpj0NeCQ07Q4AkdM/LwkmHtxoZuEp2IcDjwTlH5jZ8iYucQB4Kvi8itA8SiIppWQh0nwdCK1t0NgylZ9HfLZGjmlKtX8xD08N+v9U0oCaoUSaKVjn4h0zuwRCs3qa2ckxvvYycLGZdQhqGyMi9n0KHJmUYEUSRMlCJLbDzKwi4vVj4N+AiWa2FthAaLnYpjxKaHW1jcAfgdeBPcG+ecBfYjRNYWaDgtmGLwHuMbMNLf+RRJpHs86KtBEzO8Ld95pZV+A1YJi7f5DquETiobZQkbbzVLAAUGfgP5UoJJOoZiEiIjGpz0JERGJSshARkZiULEREJCYlCxERiUnJQkREYvr/+okPiMygYmMAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "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"
          }
        },
        {
          "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"
          }
        },
        {
          "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"
          }
        },
        {
          "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"
          }
        },
        {
          "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"
          }
        }
      ]
    }
  ]
}