House_Prices,_Advanced_Regression_Techniques-Modeling-.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<script>\n",
       "code_show=true; \n",
       "function code_toggle() {\n",
       " if (code_show){\n",
       " $('div.input').hide();\n",
       " } else {\n",
       " $('div.input').show();\n",
       " }\n",
       " code_show = !code_show\n",
       "} \n",
       "$( document ).ready(code_toggle);\n",
       "</script>\n",
       "<form action=\"javascript:code_toggle()\"><input type=\"submit\" value=\"Click here to toggle on/off the raw code.\"></form>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import HTML\n",
    "from IPython.display import Image\n",
    "\n",
    "HTML('''<script>\n",
    "code_show=true; \n",
    "function code_toggle() {\n",
    " if (code_show){\n",
    " $('div.input').hide();\n",
    " } else {\n",
    " $('div.input').show();\n",
    " }\n",
    " code_show = !code_show\n",
    "} \n",
    "$( document ).ready(code_toggle);\n",
    "</script>\n",
    "<form action=\"javascript:code_toggle()\"><input type=\"submit\" value=\"Click here to toggle on/off the raw code.\"></form>''')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "from plot_lib import plot_data, plot_model, set_default"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Running the Models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load libraries\n",
    "from sklearn.linear_model    import ElasticNet, Lasso,  BayesianRidge, LassoLarsIC\n",
    "from sklearn.ensemble        import RandomForestRegressor as rfr,  GradientBoostingRegressor\n",
    "from sklearn.kernel_ridge    import KernelRidge\n",
    "from sklearn.pipeline        import make_pipeline\n",
    "from sklearn.preprocessing   import RobustScaler\n",
    "from sklearn.base            import BaseEstimator, TransformerMixin, RegressorMixin, clone\n",
    "from sklearn.model_selection import KFold, cross_val_score, train_test_split\n",
    "from sklearn.metrics         import mean_squared_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Load data\n",
    "import pandas as pd\n",
    "import numpy  as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "test = pd.read_csv('test.csv')\n",
    "test_ID = test['Id']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "train   = pd.read_csv('train_cleaning.csv')\n",
    "test    = pd.read_csv('test_cleaning.csv')\n",
    "y_train = pd.read_csv('y_train.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_train = y_train.values.ravel()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train:  (1020, 252) \n",
      "X_test:  (438, 252) \n",
      "Y_train:  (1020,) \n",
      "Y_test:  (438,)\n"
     ]
    }
   ],
   "source": [
    "# Next we want to sample our training data to test for performance of robustness ans accuracy, before applying to the test data\n",
    "X_train, X_test, Y_train, Y_test = train_test_split(train, y_train, test_size=0.3, random_state=42)\n",
    "\n",
    "# X_train = predictor features for estimation dataset\n",
    "# X_test = predictor variables for validation dataset\n",
    "# Y_train = target variable for the estimation dataset\n",
    "# Y_test = target variable for the estimation dataset\n",
    "\n",
    "print('X_train: ', X_train.shape, '\\nX_test: ', X_test.shape, '\\nY_train: ', Y_train.shape, '\\nY_test: ', Y_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Got this from a kaggler- it's pretty handy, letting us select the number of folds and letting us shuffle the data when we run it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_folds = 10\n",
    "#this looks better, lets me use kfolds\n",
    "def rmsle_cv(model):\n",
    "    kf = KFold(n_folds, shuffle=True, random_state=1).get_n_splits(train)\n",
    "    rmse= np.sqrt(-cross_val_score(model, train, y_train, scoring=\"neg_mean_squared_error\", cv = kf))\n",
    "    return(rmse)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lasso was the first feature that I spotchecked all my changes against- It runs very quickly so it's easy to iterate off of."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "lasso = make_pipeline(RobustScaler(), Lasso(alpha =0.00099, random_state=1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Lasso score: 0.1156 (0.0132)\n"
     ]
    }
   ],
   "source": [
    "score = rmsle_cv(lasso)\n",
    "print(\"Lasso score: {:.4f} ({:.4f})\".format(score.mean(), score.std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "     steps=[('robustscaler', RobustScaler(copy=True, quantile_range=(25.0, 75.0), with_centering=True,\n",
       "       with_scaling=True)), ('lasso', Lasso(alpha=0.00099, copy_X=True, fit_intercept=True, max_iter=1000,\n",
       "   normalize=False, positive=False, precompute=False, random_state=1,\n",
       "   selection='cyclic', tol=0.0001, warm_start=False))])"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lasso.fit(X_train, Y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = list(zip(train.columns, lasso.steps[1][1].coef_))\n",
    "x.sort(key = lambda x: x[1])\n",
    "df = pd.DataFrame(x, columns=['Var', 'Coeff'])\n",
    "\n",
    "top_pos = df[\"Coeff\"][-20:]\n",
    "top_neg = df[\"Coeff\"][:20]\n",
    "top_pos_names = df[\"Var\"][-20:]\n",
    "top_neg_names = df[\"Var\"][:20]\n",
    "\n",
    "top_coefs = np.hstack([top_neg, top_pos])\n",
    "top_names = np.hstack([top_neg_names, top_pos_names])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "([<matplotlib.axis.XTick at 0x7f46602522b0>,\n",
       "  <matplotlib.axis.XTick at 0x7f46602bdc18>,\n",
       "  <matplotlib.axis.XTick at 0x7f46602bd940>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601d6550>,\n",
       "  <matplotlib.axis.XTick at 0x7f466077c7f0>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601d6ac8>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601d6e80>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601e84a8>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601e8978>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601e8e48>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601ef358>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601ef828>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601efcf8>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601efdd8>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601e88d0>,\n",
       "  <matplotlib.axis.XTick at 0x7f466077c748>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601f8518>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601f89e8>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601f8eb8>,\n",
       "  <matplotlib.axis.XTick at 0x7f46602003c8>,\n",
       "  <matplotlib.axis.XTick at 0x7f4660200898>,\n",
       "  <matplotlib.axis.XTick at 0x7f4660200d68>,\n",
       "  <matplotlib.axis.XTick at 0x7f4660208278>,\n",
       "  <matplotlib.axis.XTick at 0x7f4660200470>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601f84e0>,\n",
       "  <matplotlib.axis.XTick at 0x7f4660208518>,\n",
       "  <matplotlib.axis.XTick at 0x7f4660208978>,\n",
       "  <matplotlib.axis.XTick at 0x7f4660208e48>,\n",
       "  <matplotlib.axis.XTick at 0x7f4660192358>,\n",
       "  <matplotlib.axis.XTick at 0x7f4660192828>,\n",
       "  <matplotlib.axis.XTick at 0x7f4660192cf8>,\n",
       "  <matplotlib.axis.XTick at 0x7f466019a208>,\n",
       "  <matplotlib.axis.XTick at 0x7f4660192780>,\n",
       "  <matplotlib.axis.XTick at 0x7f4660208a20>,\n",
       "  <matplotlib.axis.XTick at 0x7f4660200710>,\n",
       "  <matplotlib.axis.XTick at 0x7f466019a978>,\n",
       "  <matplotlib.axis.XTick at 0x7f466019ae48>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601a2358>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601a2828>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601a2cf8>],\n",
       " <a list of 40 Text xticklabel objects>)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (15, 5))\n",
    "colors = ['red' if c < 0 else 'blue' for c in top_coefs]\n",
    "plt.bar(np.arange(40), top_coefs, color = colors)\n",
    "feature_names = np.array(train.columns)\n",
    "plt.xticks(np.arange(40), top_names, rotation = 60, ha = 'right')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "switched away from boxcox to log1p and instantly improved by over 10%"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "ENet ended up being our best individual performer before we switched to stacking.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "ENet = make_pipeline(RobustScaler(), ElasticNet(alpha=0.0005, l1_ratio=.9, random_state=1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "#ok so alpha here is the tuning parameter- L1 ratio is the % of penalty assigned to L1 or L2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ElasticNet score: 0.1130 (0.0135)\n"
     ]
    }
   ],
   "source": [
    "score = rmsle_cv(ENet)\n",
    "print(\"ElasticNet score: {:.4f} ({:.4f})\".format(score.mean(), score.std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "     steps=[('robustscaler', RobustScaler(copy=True, quantile_range=(25.0, 75.0), with_centering=True,\n",
       "       with_scaling=True)), ('elasticnet', ElasticNet(alpha=0.0005, copy_X=True, fit_intercept=True, l1_ratio=0.9,\n",
       "      max_iter=1000, normalize=False, positive=False, precompute=False,\n",
       "      random_state=1, selection='cyclic', tol=0.0001, warm_start=False))])"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ENet.fit(X_train, Y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = list(zip(train.columns, ENet.steps[1][1].coef_))\n",
    "x.sort(key = lambda x: x[1])\n",
    "df = pd.DataFrame(x, columns=['Var', 'Coeff'])\n",
    "\n",
    "top_pos = df[\"Coeff\"][-20:]\n",
    "top_neg = df[\"Coeff\"][:20]\n",
    "top_pos_names = df[\"Var\"][-20:]\n",
    "top_neg_names = df[\"Var\"][:20]\n",
    "\n",
    "top_coefs = np.hstack([top_neg, top_pos])\n",
    "top_names = np.hstack([top_neg_names, top_pos_names])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "([<matplotlib.axis.XTick at 0x7f466012db70>,\n",
       "  <matplotlib.axis.XTick at 0x7f466012d3c8>,\n",
       "  <matplotlib.axis.XTick at 0x7f46601efe48>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f280cf8>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f2221d0>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f2226a0>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f222b70>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f22a0f0>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f222668>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f22a390>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f22a860>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f22ad30>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f232240>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f232710>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f232be0>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f23b160>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f23b5c0>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f2326d8>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f22a908>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f23b940>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f23bd30>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f244240>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f244710>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f244be0>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f1cd160>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f1cd5c0>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f1cda90>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f244a58>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f232208>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f1cd198>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f1d5240>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f1d5710>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f1d5be0>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f1dd160>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f1dd5c0>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f1dda90>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f1ddf60>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f1dd588>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f1cd0b8>,\n",
       "  <matplotlib.axis.XTick at 0x7f465f1e62b0>],\n",
       " <a list of 40 Text xticklabel objects>)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (15, 5))\n",
    "colors = ['red' if c < 0 else 'blue' for c in top_coefs]\n",
    "plt.bar(np.arange(40), top_coefs, color = colors)\n",
    "feature_names = np.array(train.columns)\n",
    "plt.xticks(np.arange(40), top_names, rotation = 60, ha = 'right')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- .9 l1_ratio gives score: 0.1098 (0.0145)\n",
    "- .5  gives ElasticNet score: 0.1108 (0.0144)\n",
    "- .7 gives .1099 (0.0146)\n",
    "- alpha=0.0003, l1_ratio=.9 gives score: 0.1107 (0.0145) # while better here it's worse in my actual score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "GBoost = GradientBoostingRegressor(n_estimators=3000, learning_rate=0.05,\n",
    "                                   max_depth=4, max_features='sqrt',\n",
    "                                   min_samples_leaf=15, min_samples_split=10, \n",
    "                                   loss='huber', random_state =5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gradient Boosting score: 0.1147 (0.0156)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "score = rmsle_cv(GBoost)\n",
    "print(\"Gradient Boosting score: {:.4f} ({:.4f})\\n\".format(score.mean(), score.std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GradientBoostingRegressor(alpha=0.9, criterion='friedman_mse', init=None,\n",
       "             learning_rate=0.05, loss='huber', max_depth=4,\n",
       "             max_features='sqrt', max_leaf_nodes=None,\n",
       "             min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "             min_samples_leaf=15, min_samples_split=10,\n",
       "             min_weight_fraction_leaf=0.0, n_estimators=3000,\n",
       "             n_iter_no_change=None, presort='auto', random_state=5,\n",
       "             subsample=1.0, tol=0.0001, validation_fraction=0.1, verbose=0,\n",
       "             warm_start=False)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GBoost.fit(X_train, Y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "index = np.argsort(GBoost.feature_importances_)\n",
    "features = GBoost.feature_importances_[index]\n",
    "features = features[-40:]\n",
    "names = np.array(X_train)[index]\n",
    "names = names[-40:0]\n",
    "\n",
    "# # plot as bar chart\n",
    "# plt.barh(np.arange(len(names)), features)\n",
    "# plt.yticks(np.arange(len(names)) + .25, names)\n",
    "# _ = plt.xlabel('Relative importance')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAqIAAAJQCAYAAABVWwbhAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4xLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvDW2N/gAAIABJREFUeJzt3XuUXnV97/H3x6CBcAlUKEYuTtVQ5Rol4gU8IkWrxaNwQKFVC8WW2uKltF6olxbL0UOtFqzosdGFYI8KRYSinCPITdCKuUBIAoIIhFa0tYqNXEXC9/zx7NGHNJOZZCbzm2fm/VrrWbOf3977t7/7t8j48bf3np2qQpIkSZpsj2tdgCRJkmYmg6gkSZKaMIhKkiSpCYOoJEmSmjCISpIkqQmDqCRJkpowiEqSJKkJg6gkSZKaMIhKkiSpiS1aF6Cx2XHHHWtoaKh1GZIkSaNatmzZj6pqp9G2M4gOiKGhIZYuXdq6DEmSpFEluWss23lpXpIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTWzRugCNzcq71zB08iWty5AkSQNs9WmHtS7hMZwRlSRJUhMGUUmSJDVhEJUkSVIT0yqIJtk5yeeS3JFkWZJvJjliPdsNJVm1nva/SnLoGI7zrCSV5DcnqnZJkqSZZtoE0SQBLgKuqaqnVtX+wDHArutsN+IDWlX1F1V1+RgO99vA17uf660lybQZW0mSpM1hOoWlQ4CHq+oTww1VdVdVfTTJcUnOT/Il4LKROkhydpKjkrw8yT/2tR/c7TsceI8CjgNemmTLrn0oybeTfBy4HtgtyUu7Wdnru+Nv0237F0mWJFmVZFHXpyRJ0owynYLoXvQC4EieDxxbVYeMoa+vAs9LsnX3/WjgvG75QODOqroduBr4rb79fh34TFU9C7gfeA9waFU9G1gK/Gm33ZlV9Zyq2hvYCnjF+opIckKSpUmWrn1gzRjKliRJGhzTKYg+RpKPJbkxyZKu6atVdc9Y9q2qR4CvAP+9u5R/GPBP3erfBs7tls/lsZfn76qq67rl5wF7At9Ishw4FnhKt+7FSb6VZCW9mdy9RqhjUVUtrKqFs+bMHUvpkiRJA2M6/UH7m4Ajh79U1YlJdqQ3Ewm9GcqNcR5wInAPsKSq7k0yqzvGK5O8GwjwxCTbrucYoRd+H3MfaXcp/+PAwqr61ySnAFtuZG2SJEkDbzrNiF4JbJnkj/ra5oyjv6uBZwN/wC8vyx8K3FhVu1XVUFU9BbgAOHw9+18HHJjk6QBJ5iTZg1+Gzh9194weNY4aJUmSBta0CaJVVfQC4YuS3JlkMXAO8M4Rdvn1JN/r+7x6nf7WAl8GXt79hN5l+AvX6ecC4HfWU89/0Hug6fNJVtALps+oqv8EPgmspPeU/5J195UkSZoJ0stvmupmz5tf8449o3UZkiRpgE3Wu+aTLKuqhaNtN21mRCVJkjRYDKKSJElqwiAqSZKkJgyikiRJasIgKkmSpCYMopIkSWrCICpJkqQmDKKSJElqwiAqSZKkJgyikiRJasIgKkmSpCYMopIkSWrCICpJkqQmDKKSJElqwiAqSZKkJgyikiRJasIgKkmSpCYMopIkSWrCICpJkqQmDKKSJElqwiAqSZKkJgyikiRJasIgKkmSpCYMopIkSWrCICpJkqQmDKKSJElqwiAqSZKkJrZoXYDGZp9d5rL0tMNalyFJkjRhnBGVJElSEwZRSZIkNWEQlSRJUhMGUUmSJDVhEJUkSVITBlFJkiQ1YRCVJElSEwZRSZIkNWEQlSRJUhO+WWlArLx7DUMnX9K6DEnaKKt9I5ykDXBGVJIkSU0YRCVJktSEQVSSJElNDEwQTbJzks8luSPJsiTfTHJEw3penmRpkm8nuSXJh1rVIkmSNIgGIogmCXARcE1VPbWq9geOAXYd4/6zJrievYEzgddV1TOBvYE7NmJ/HxKTJEkz3kAEUeAQ4OGq+sRwQ1XdVVUfTTKU5Nok13efFwAkOTjJVUk+B6zs2i7qZlNvSnLCcF9J3pDkO0muTvLJJGd27TsluSDJku5zYLfLO4D3V9UtXS2PVNXHu33+e5JvJbkhyeVJdu7aT0myKMllwGeS7JVkcZLlSVYkmb/ZR1GSJGkKGZSZub2A60dY90PgJVX1UBfmPg8s7NYdAOxdVXd234+vqnuSbAUsSXIBMBt4L/Bs4F7gSuDGbvuPAKdX1deT7A5cCgzPgH54hHq+DjyvqirJ79MLrX/WrdsfOKiqHkzyUeAjVfXZJE8AJnTWVpIkaaoblCD6GEk+BhwEPAwcCpyZZAGwFtijb9PFfSEU4C1995XuBswHngR8raru6fo+v6+PQ4E9e3cGALBdkm1HKW9X4Lwk84AnAP3Hv7iqHuyWvwm8O8muwBer6rb1nOcJwAkAs7bbaZTDSpIkDZZBuTR/E70ZSwCq6kTgN4CdgJOAfwf2ozcT+oS+/e4fXkhyML1g+fyq2g+4AdgSCCN7XLf9gu6zS1Xd29Wz/wj7fBQ4s6r2Af6wO8Z/qaeqPge8EngQuDTJIet2VFWLqmphVS2cNWfuBsqUJEkaPIMSRK8EtkzyR31tc7qfc4EfVNWjwOsZ+RL3XOAnVfVAkmcAz+vaFwMvSrJD9xDRkX37XAa8afhLN+sK8DfAu5Ls0bU/Lsmf9h3n7m752JFOKMlTgTuq6u+Ai4F9R9pWkiRpOhqIIFpVBRxOLzDemWQxcA7wTuDjwLFJrqN3Sf3+Ebr5CrBFkhXAqcB1Xd93Ax8AvgVcDtwMrOn2eQuwsHuY6Gbgjd0+K4A/AT6f5NvAKmBet88pwPlJrgV+tIHTOhpYlWQ58AzgM2MfEUmSpMGXXsab2ZJsU1X3dTOiFwJnVdWFrevqN3ve/Jp37Bmty5CkjeK75qWZKcmyqlo42nYDMSM6CU7pZiZX0Xu46KLG9UiSJE17A/nU/ESrqre1rkGSJGmmcUZUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1MQWrQvQ2Oyzy1yWnnZY6zIkSZImjDOikiRJasIgKkmSpCYMopIkSWrCICpJkqQmDKKSJElqwiAqSZKkJgyikiRJasIgKkmSpCYMopIkSWrCNysNiJV3r2Ho5EtalyFpHFb7djRJegxnRCVJktSEQVSSJElNGEQlSZLUxMAE0SS7JvmnJLcluT3JR5I8YTMf877u51CSVX3tByVZnOSWJLcmOXEijiNJkjSTDEQQTRLgi8BFVTUf2APYBnj/OPvd6Ie1kjwJ+Bzwxqp6BnAgcHySI8ZTiyRJ0kwzEEEUOAR4qKo+DVBVa4GT6AXAJUn2Gt4wydVJ9k+ydZKzuvU3JHlVt/64JOcn+RJwWZJtklyR5PokK4e324ATgbOr6vqulh8B7wDe3vV/dpKj+uoZnlXd2ONIkiRNa4Py55v2Apb1N1TVT5P8C/Bl4DXAXyaZBzy5qpYl+QBwZVUdn2R7YHGSy7vdnw/sW1X3dLOiR3T97Qhcl+TiqqoN1HLOOm1LgT1HOYeHNvI4JDkBOAFg1nY7jdK9JEnSYBmUGdEA6wtsAa4GXt19fw1wfrf8UuDkJMu7bbYEdu/WfbWq7unr4wNJVgCXA7sAO29CLWM5h405DlW1qKoWVtXCWXPmbsIhJUmSpq5BmRG9CTiyvyHJdsBuwBLgx0n2BY4G/nB4E+DIqrp1nf2eC9zf1/RaYCdg/6r6eZLV9ELrhmpZCFzc17Y/vVlRgEfoAn53b+vwA1UbexxJkqRpbVBmRK8A5iT5XYAks4AP07tX8wHgXHr3ac6tqpXdPpcCb+7CIEmeNULfc4EfduHwxcBTRqnlY8BxSRZ0/T6R3kNTp3brV9MLpgCvAh6/iceRJEma1gYiiHb3UR4BvDrJbcB36N1z+a5uky8AxwD/2LfbqfRC4IruTy+dyvp9FliYZCm9WctbRqnlB8DrgEVJbgW+D/xdVX2t2+STwIuSLAb6Z1836jiSJEnTXTbwrIzGoPsbom8E/ltV/WRzHWf2vPk179gzNlf3kiaB75qXNFMkWVZVC0fbbiBmRKeyqvpYVe2zOUOoJEnSdGQQlSRJUhMGUUmSJDVhEJUkSVITBlFJkiQ1YRCVJElSEwZRSZIkNWEQlSRJUhMGUUmSJDVhEJUkSVITBlFJkiQ1YRCVJElSEwZRSZIkNWEQlSRJUhMGUUmSJDVhEJUkSVITBlFJkiQ1YRCVJElSEwZRSZIkNWEQlSRJUhMGUUmSJDVhEJUkSVITBlFJkiQ1YRCVJElSEwZRSZIkNWEQlSRJUhMGUUmSJDWxResCNDb77DKXpacd1roMSZKkCeOMqCRJkpowiEqSJKkJg6gkSZKaMIhKkiSpCYOoJEmSmjCISpIkqQmDqCRJkpowiEqSJKkJg6gkSZKa8M1KA2Ll3WsYOvmS1mVIza32DWOSNG04IypJkqQmDKKSJElqwiAqSZKkJqZtEE3P15O8vK/tNUm+MgF9/58kdyZZnuSWJO8Zwz5HJHl7t/w/k/xJt3x8kieNtyZJkqRBM20fVqqqSvJG4PwkVwGzgPcDLxtPv0mGx+ykqrooyVbALUnOqap/3UA9F46w6njgeuDfxlOXJEnSoJm2M6IAVbUK+BLwTuAvgc9U1e1Jjk2yuJvR/HiSxwEkWZRkaZKbkvzFcD9JvpfkvUm+ARyxzmG2Agp4oG/b7bvl5yW5vFv+/SRn9O+Y5GhgAXBeV8sTNsc4SJIkTUXTOoh23gf8DvBy4INJ9qYXJl9QVQvozQof0217clUtBPYDXpJkz75+7q+qA6vq/O776UmWA/9KL+D+eGMLq6rzgOXA0VW1oKoe3pQTlCRJGkTT9tL8sKq6P8l5wH1V9bMkhwLPAZYmgd6M5vAl9d9O8gZ64/JkYE/g5m7deet0PXxpflvgqiRfrqrFE1l7khOAEwBmbbfTRHYtSZLU3LQPop1Huw9AgLOq6r39GySZD7wVOKCq/jPJ/wG27Nvk/vV1XFX3JvkacBCwGHiEX840b7m+fcaqqhYBiwBmz5tf4+lLkiRpqpkJl+bXdTnwmiQ7AiR5YpLdge2Ae4GfJpkH/OZYOkvyeOAA4PauaTWwf7d85Bi6uBfYdszVS5IkTRMzZUb0F6pqZZL3AZd3Dyn9HHgjsJTeZfhVwB3AN0bp6vQkpwCzgUuBi7v2U4BPJvk3ejOko/k08KkkD9KbjfU+UUmSNCOkyiu+g2D2vPk179gzRt9QmuZ817wkTX1JlnUPgG/QTLw0L0mSpCnAICpJkqQmDKKSJElqwiAqSZKkJgyikiRJasIgKkmSpCYMopIkSWrCICpJkqQmDKKSJElqwiAqSZKkJgyikiRJasIgKkmSpCYMopIkSWrCICpJkqQmDKKSJElqwiAqSZKkJgyikiRJasIgKkmSpCYMopIkSWrCICpJkqQmDKKSJElqwiAqSZKkJgyikiRJasIgKkmSpCYMopIkSWrCICpJkqQmDKKSJElqwiAqSZKkJrZoXYDGZp9d5rL0tMNalyFJkjRhnBGVJElSEwZRSZIkNWEQlSRJUhMGUUmSJDVhEJUkSVITBlFJkiQ1YRCVJElSEwZRSZIkNWEQlSRJUhO+WWlArLx7DUMnX9K6DKmZ1b5ZTJKmHWdEJUmS1IRBVJIkSU0YRCVJktTEjAuiSc5K8sMkq0bZ7uAkL+j7fkqSu5Ms7z6nde1XJ1k4Qh+vSHJDkhuT3JzkDzfUlyRJ0kwyEx9WOhs4E/jMKNsdDNwH/HNf2+lV9aGxHCTJbGARcEBVfa/7PrQpfUmSJE1HM25GtKquAe7pb0vylm7GckWSc5MMAW8ETupmLF84lr6T3Jfkr5J8C3guvaD/4+64P6uqWyfyXCRJkgbZjAuiIzgZeFZV7Qu8sapWA5+gN2u5oKqu7bY7qe9y+m+up5+tgVVV9dwu8F4M3JXk80lem6R/vEfriyQnJFmaZOnaB9ZM2MlKkiRNBQbRnhXAZ5O8DnhkA9sNB9MFVXXpetavBS4Y/lJVvw/8BrAYeBtw1kb0RVUtqqqFVbVw1py5G3tOkiRJU5pBtOcw4GPA/sCyJJt67+xDVbW2v6GqVlbV6cBLgCPHV6YkSdL0MeODaHe5fLequgp4B7A9sA1wL7DtOPrdJsnBfU0LgLvGUaokSdK0MuOemk/yeXpPxO+Y5HvAqcDrk8wFQu+S+X8m+RLwhSSvAt68KYcC3pHk74EHgfuB4ybgFCRJkqaFVFXrGjQGs+fNr3nHntG6DKkZ3zUvSYMjybKqWu/fWe834y/NS5IkqQ2DqCRJkpowiEqSJKkJg6gkSZKaMIhKkiSpCYOoJEmSmjCISpIkqQmDqCRJkpowiEqSJKkJg6gkSZKaMIhKkiSpCYOoJEmSmjCISpIkqQmDqCRJkpowiEqSJKkJg6gkSZKaMIhKkiSpCYOoJEmSmjCISpIkqQmDqCRJkpowiEqSJKkJg6gkSZKaMIhKkiSpCYOoJEmSmjCISpIkqQmDqCRJkpowiEqSJKmJLVoXoLHZZ5e5LD3tsNZlSJIkTRhnRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhG9WGhAr717D0MmXtC5DmhCrfUuYJAlnRCVJktSIQVSSJElNGEQlSZLUxLQMokl2TfJPSW5LcnuSjyR5wmY+5n3dz6Ekq7rlOUk+m2RlklVJvp5km27d2iTL+z5Dm7M+SZKkqWbaBdEkAb4IXFRV84E9gG2A94+z3015sOutwL9X1T5VtTfwBuDn3boHq2pB32f1eOqTJEkaNNMuiAKHAA9V1acBqmotcBJwfJIlSfYa3jDJ1Un2T7J1krO69TckeVW3/rgk5yf5EnBZkm2SXJHk+m6W81Wj1DIPuHv4S1XdWlU/m+gTliRJGkTT8c837QUs62+oqp8m+Rfgy8BrgL9MMg94clUtS/IB4MqqOj7J9sDiJJd3uz8f2Leq7ulmRY/o+tsRuC7JxVVVI9RyFr0AexRwBXBOVd3WrdsqyfJu+c6qOmKCzl+SJGkgTMcZ0QDrC4YBrgZe3X1/DXB+t/xS4OQuGF4NbAns3q37alXd09fHB5KsAC4HdgF2HqmQqloOPBX4G+BXgCVJntmt7r80v94QmuSEJEuTLF37wJoNn7UkSdKAmY4zojcBR/Y3JNkO2A1YAvw4yb7A0cAfDm8CHFlVt66z33OB+/uaXgvsBOxfVT9PsppeaB1RVd1H757VLyZ5FPgt4NtjOZGqWgQsApg9b/5Is66SJEkDaTrOiF4BzEnyuwBJZgEfBs6uqgeAc4F3AHOramW3z6XAm7sHnUjyrBH6ngv8sAuhLwaesqFCkhyYZIdu+QnAnsBd4zo7SZKkaWLaBdHufs0jgFcnuQ34DvAQ8K5uky8AxwD/2LfbqcDjgRXdn146dYTuPwssTLKU3uzoLaOU8zTga0lWAjcAS4ELNvqkJEmSpqGM/JyNppLZ8+bXvGPPaF2GNCF817wkTW9JllXVwtG2m3YzopIkSRoMBlFJkiQ1YRCVJElSEwZRSZIkNWEQlSRJUhMGUUmSJDVhEJUkSVITBlFJkiQ1YRCVJElSEwZRSZIkNWEQlSRJUhMGUUmSJDVhEJUkSVITBlFJkiQ1YRCVJElSEwZRSZIkNWEQlSRJUhMGUUmSJDVhEJUkSVITBlFJkiQ1YRCVJElSEwZRSZIkNWEQlSRJUhMGUUmSJDVhEJUkSVITBlFJkiQ1YRCVJElSE1u0LkBjs88uc1l62mGty5AkSZowzohKkiSpCYOoJEmSmjCISpIkqQmDqCRJkpowiEqSJKkJg6gkSZKaMIhKkiSpCYOoJEmSmjCISpIkqQnfrDQgVt69hqGTL2ldhrTJVvtmMEnSOpwRlSRJUhMGUUmSJDVhEJUkSVITBlFJkiQ1MWoQTbI2yfK+z9DmLCjJfaOs3z7JH/d9f3KSL0xwDauTrExyY5LLkjypa98myd8nuT3JTUmuSfLciTy2JEnSTDGWGdEHq2pB32f15i5qFNsDvwiiVfX9qjpqMxznxVW1H7AUeFfX9ingHmB+Ve0FHAfsuBmOLUmSNO1t0qX5JFsm+XQ3a3hDkhd37cclObNvuy8nObhbvi/J+7tZxuuS7Ny1/1qSbyZZkuTUvn23SXJFkuu747yqW3Ua8LRudvZvkgwlWTWGur6Y5CtJbkvywY043WuApyd5GvBc4D1V9ShAVd1RVZd0x/jTJKu6z590bUNJvp3kk90M6mVJturWPT3J5d14XN/1L0mSNGOMJYhu1XdZ/sKu7USAqtoH+G3gnCRbjtLP1sB13SzjNcAfdO0fAf53VT0H+Le+7R8CjqiqZwMvBj6cJMDJwO3d7Ozb1znGhupaABwN7AMcnWS3MZw7wCuAlcBewPKqWrvuBkn2B36PXlB9HvAHSZ7VrZ4PfKybQf1P4Miu/bNd+37AC4AfrKffE5IsTbJ07QNrxliuJEnSYNjYS/NHdG0HAf8AUFW3AHcBe4zSz8PAl7vlZcBQt3wg8Plu+R/6tg/wgSQrgMuBXYCdRznGhuq6oqrWVNVDwM3AU0bp66oky4HtgP81huNeWFX3V9V9wBeBF3br7qyq5d3yMmAoybbALlV1YVfrQ1X1wLqdVtWiqlpYVQtnzZk7SgmSJEmDZVPfrJQR2h/hseG2f5b051VV3fLadY5d/FevBXYC9q+qnydZvU5/G1MXwM/6ltc9/vq8uKp+9IuOk5uA/ZI8bvjS/CYed6tRtpckSZoRNvXPN11DLyiSZA9gd+BWYDWwIMnjukvfB4yhr28Ax3TLr+1rnwv8sAuhL+aXM5j3AttuZF3jVlW303tw6X3dLQIkmd/du3oNcHiSOUm2Bo4Art1AXz8Fvpfk8K6f2UnmTESdkiRJg2JTg+jHgVlJVgLnAcdV1c/ohco76d1T+SHg+jH09VbgxCRL6IXPYZ8FFiZZSi9c3gJQVT8GvtE9FPQ3Y6xrovw+8CTgu90xPgl8v6quB84GFgPfAj5VVTeM0tfrgbd0tx78c9evJEnSjJFfXi3XVDZ73vyad+wZrcuQNtnq0w5rXYIkaZIkWVZVC0fbzjcrSZIkqYlNfVhpWkjyLWD2Os2vr6qVLeqRJEmaSWZ0EK0qX88pSZLUiJfmJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTWzRugCNzT67zGXpaYe1LkOSJGnCOCMqSZKkJgyikiRJasIgKkmSpCYMopIkSWrCICpJkqQmDKKSJElqwiAqSZKkJgyikiRJasIgKkmSpCZ8s9KAWHn3GoZOvqR1GZphVvs2L0nSZuSMqCRJkpowiEqSJKkJg6gkSZKamDJBNMmuSf4pyW1Jbk/ykSRPmOBjnJLk7iTLk6xK8soJ6ve+Edp/PcnV3fG+nWRR135wkjVd+/Ikl09EHZIkSYNkSgTRJAG+CFxUVfOBPYBtgPdvhsOdXlULgFcDZyUZ0xgk2ZQHu/5u+HhV9Uzgo33rru3aF1TVoZvQtyRJ0kCbEkEUOAR4qKo+DVBVa4GTgOOT/HE3U/qVJLcm+cvhnZK8Lsniblbx75PM6trvS/L+JDcmuS7JzusesKq+DTwC7JjkKUmuSLKi+7l718/ZSf42yVXAXyfZJsmnk6zstj2yr5b1HW8e8L2+Y66c6IGTJEkaVFMliO4FLOtvqKqfAv9C709MHQC8FlgAvDrJwiTPBI4GDuxmONd22wBsDVxXVfsB1wB/sO4BkzwXeBT4D+BM4DNVtS/wWXozmcP2AA6tqj8D3gusqap9um2vHOV4pwNXJvl/SU5Ksn1fvy/suzT/7vUNSpITkixNsnTtA2s2MHySJEmDZ6r8HdEAtYH2r1bVjwGSfBE4iN5s5v7Akt6VfbYCftjt9zDw5W55GfCSvj5PSvI64F7g6KqqJM8H/ke3/h+AD/Ztf343QwtwKHDM8Iqq+smGjldVn05yKfAy4FXAHybZr9vu2qp6xYYGpaoWAYsAZs+bv77xkSRJGlhTJYjeBBzZ35BkO2A3ejOd64awohdSz6mqP19Pfz+vquF91vLY8zy9qj40Sj39x7u/v6z11LLB41XV94Gz6N2PugrYe5RjS5IkzQhT5dL8FcCcJL8L0N3r+WHgbOAB4CVJfiXJVsDhwDe6fY5K8qvdPr+S5CmbePx/5pczna8Fvj7CdpcBbxr+kmSHDXWa5GVJHt8tPwl4InD3JtYoSZI0rUyJINrNJh5B7/7P24DvAA8B7+o2+Tq9S+bLgQuqamlV3Qy8B7gsyQrgq/QeDtoUbwF+r+vn9cBbR9jufwI7dH/66UbgxaP0+1JgeNtLgbdX1b9tYo2SJEnTSn55RXlqSnIcsLCq3jTattPZ7Hnza96xZ7QuQzOM75qXJG2KJMuqauFo202JGVFJkiTNPFPlYaURVdXZ9O4VlSRJ0jTijKgkSZKaMIhKkiSpCYOoJEmSmjCISpIkqQmDqCRJkpowiEqSJKkJg6gkSZKaMIhKkiSpCYOoJEmSmjCISpIkqQmDqCRJkpowiEqSJKkJg6gkSZKaMIhKkiSpCYOoJElm29J0AAARtElEQVSSmjCISpIkqQmDqCRJkpowiEqSJKkJg6gkSZKaMIhKkiSpCYOoJEmSmjCISpIkqQmDqCRJkpowiEqSJKkJg6gkSZKa2KJ1ARqbfXaZy9LTDmtdhiRJ0oRxRlSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTfhmpQGx8u41DJ18Sesy1NBq36wlSZpmnBGVJElSEwZRSZIkNWEQlSRJUhMDEUST7Jzkc0nuSLIsyTeTHNG4pn9K8s2WNUiSJA2yKR9EkwS4CLimqp5aVfsDxwC7jnH/WZuhpu2BZwPbJ/m1EbbxQTBJkqQNmPJBFDgEeLiqPjHcUFV3VdVHkwwluTbJ9d3nBQBJDk5yVZLPASu7tou62dSbkpww3FeSNyT5TpKrk3wyyZld+05JLkiypPsc2FfTkcCXgHPpheLhvs5O8rdJrgL+OsnWSc7q9r8hyau67dZbtyRJ0kwyCLN2ewHXj7Duh8BLquqhJPOBzwMLu3UHAHtX1Z3d9+Or6p4kWwFLklwAzAbeS292817gSuDGbvuPAKdX1deT7A5cCjyzW/fbwPuAfwe+APyvvpr2AA6tqrVJPgBcWVXHd7Ooi5NcPkrdkiRJM8IgBNHHSPIx4CDgYeBQ4MwkC4C19ELgsMV9IRTgLX33le4GzAeeBHytqu7p+j6/r49DgT17dwYAsF2SbYE5wNOBr1dVJXkkyd5Vtarb7vyqWtstvxR4ZZK3dd+3BHYHvr+BuvvP9QTgBIBZ2+00tgGSJEkaEIMQRG+idykcgKo6McmOwFLgJHqzkvvRu83gob797h9eSHIwvWD5/Kp6IMnV9EJhGNnjuu0f7G9M8nvADsCdXUjdjt7l+fese9yu/yOr6tZ1+jhlA3X/QlUtAhYBzJ43vzZQqyRJ0sAZhHtErwS2TPJHfW1zup9zgR9U1aPA64GRHkyaC/ykC6HPAJ7XtS8GXpRkh+7hoiP79rkMeNPwl272EnqX5V9WVUNVNQQMPzy1PpcCb+4euCLJszaybkmSpGlrygfRqirgcHqB8c4ki4FzgHcCHweOTXIdvcvb94/QzVeALZKsAE4Fruv6vhv4APAt4HLgZmBNt89bgIVJViS5GXhjkiF6l9av66vvTuCnSZ67nuOeCjweWJFkVfedjahbkiRp2kov581cSbapqvu6GdELgbOq6sLWda1r9rz5Ne/YM1qXoYZ817wkaVAkWVZVoz6IPeVnRCfBKUmWA6uAO+n9zVJJkiRtZoPwsNJmVVVvG30rSZIkTTRnRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNbNG6AI3NPrvMZelph7UuQ5IkacI4IypJkqQmDKKSJElqwiAqSZKkJgyikiRJasIgKkmSpCYMopIkSWrCICpJkqQmDKKSJElqwiAqSZKkJnyz0oBYefcahk6+pHUZmkSrfZOWJGmac0ZUkiRJTRhEJUmS1IRBVJIkSU1M+SCaZOckn0tyR5JlSb6Z5IgGdeyV5DtJtupruyTJMevZ9uAka5IsT7IiyeVJfrVbd1ySM7vlw5PsOXlnIUmSNHVM6SCaJMBFwDVV9dSq2h84Bth1jPvPmqhaquom4IvAu7u+DwceX1XnrnPM4QfArq2qBVW1L7AEOHE93R4OGEQlSdKMNKWDKHAI8HBVfWK4oaruqqqPJhlKcm2S67vPC+AXs5FXJfkcsLJru6ibTb0pyQnDfSV5QzfLeXWST/bNVO6U5IIkS7rPgd0ufwW8OskC4DS6cJnklCSLklwGfKb/BLowvS3wk3XaXwC8Evibbub0aRM4bpIkSVPeVP/zTXsB14+w7ofAS6rqoSTzgc8DC7t1BwB7V9Wd3ffjq+qe7rL6kiQXALOB9wLPBu4FrgRu7Lb/CHB6VX09ye7ApcAzq+qBJG8DrgH+tqpu66tnf+CgqnowycHAC5MsB54I3A+8q7/4qvrnJBcDX66qL6zvBLvQfALArO122vBISZIkDZipHkQfI8nHgIOAh4FDgTO72cm1wB59my7uC6EAb+m7r3Q3YD7wJOBrVXVP1/f5fX0cCuzZm8wEYLsk21bVvVX1pST/CXx8nfIurqoH+75fW1Wv6Pp+J/BB4I0bc75VtQhYBDB73vzamH0lSZKmuqkeRG8Cjhz+UlUnJtkRWAqcBPw7sB+9Wwwe6tvv/uGFbnbyUOD53Yzm1cCWQBjZ47rtHxxh/aPdp9/969uwczFwwQbWS5IkzThT/R7RK4Etk/xRX9uc7udc4AdV9SjwemCkB5PmAj/pQugzgOd17YuBFyXZoXvA6Mi+fS4D3jT8pZt1HY+DgNvX034vvftHJUmSZpwpHUSrqug9Wf6iJHcmWQycA7yT3qXxY5NcR++S+kgzkl8BtkiyAjgVuK7r+27gA8C3gMuBm4E13T5vARZ2f3rpZjbyknrnhd1DSDfSC8p/tp5tzgXenuQGH1aSJEkzTXpZb2ZKsk1V3dfNiF4InFVVF7aua31mz5tf8449o3UZmkS+a16SNKiSLKuqhaNtN6VnRCfBKd2T7auAO+n9zVJJkiRNgqn+sNJmVVVva12DJEnSTDXTZ0QlSZLUiEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTRhEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUxBatC9DY7LPLXJaedljrMiRJkiaMM6KSJElqwiAqSZKkJgyikiRJasIgKkmSpCYMopIkSWrCICpJkqQmDKKSJElqwiAqSZKkJgyikiRJasIgKkmSpCYMopIkSWrCICpJkqQmDKKSJElqwiAqSZKkJgyikiRJasIgKkmSpCYMopIkSWrCICpJkqQmDKKSJElqwiAqSZKkJgyikiRJasIgKkmSpCYMopIkSWrCICpJkqQmUlWta9AYJLkXuLV1HQNgR+BHrYsYEI7V2DhOY+M4jZ1jNTaO09hM1XF6SlXtNNpGW0xGJZoQt1bVwtZFTHVJljpOY+NYjY3jNDaO09g5VmPjOI3NoI+Tl+YlSZLUhEFUkiRJTRhEB8ei1gUMCMdp7ByrsXGcxsZxGjvHamwcp7EZ6HHyYSVJkiQ14YyoJEmSmjCITgFJXpbk1iTfTXLyetbPTnJet/5bSYb61v15135rkt+czLon26aOU5KXJFmWZGX385DJrn0yjee/p2797knuS/K2yaq5lXH+29s3yTeT3NT9t7XlZNY+mcbxb+/xSc7pxufbSf58smufTGMYp/+W5PokjyQ5ap11xya5rfscO3lVT75NHackC/r+za1IcvTkVj75xvPfVLd+uyR3JzlzcireBFXlp+EHmAXcDjwVeAJwI7DnOtv8MfCJbvkY4Lxuec9u+9nAr3X9zGp9TlNwnJ4FPLlb3hu4u/X5TMVx6lt/AXA+8LbW5zNVx4ren75bAezXfX+i//bWO06/A5zbLc8BVgNDrc+p4TgNAfsCnwGO6mv/FeCO7ucO3fIOrc9pCo7THsD8bvnJwA+A7Vuf01Qcq771HwE+B5zZ+nxG+jgj2t4BwHer6o6qehg4F3jVOtu8CjinW/4C8BtJ0rWfW1U/q6o7ge92/U1HmzxOVXVDVX2/a78J2DLJ7EmpevKN578nkhxO738Eb5qkelsaz1i9FFhRVTcCVNWPq2rtJNU92cYzTgVsnWQLYCvgYeCnk1P2pBt1nKpqdVWtAB5dZ9/fBL5aVfdU1U+ArwIvm4yiG9jkcaqq71TVbd3y94EfAqP+wfQBNp7/pkiyP7AzcNlkFLupDKLt7QL8a9/373Vt692mqh4B1tCbgRnLvtPFeMap35HADVX1s81UZ2ubPE5JtgbeCbxvEuqcCsbz39QeQCW5tLss9o5JqLeV8YzTF4D76c1c/Qvwoaq6Z3MX3Mh4fh/7u3wjJTmA3izh7RNU11S0yWOV5HHAh4G3b4a6JpRvVmov62lb908ZjLTNWPadLsYzTr2VyV7AX9ObzZquxjNO7wNOr6r7ugnS6W48Y7UFcBDwHOAB4Ioky6rqioktcUoYzzgdAKyldxl1B+DaJJdX1R0TW+KUMJ7fx/4u35gOknnAPwDHVtV/mQmcRsYzVn8M/N+q+tep/vvcGdH2vgfs1vd9V+D7I23TXeKaC9wzxn2ni/GME0l2BS4EfreqpvP/gx7POD0X+GCS1cCfAO9K8qbNXXBD4/2397Wq+lFVPQD8X+DZm73iNsYzTr8DfKWqfl5VPwS+AQzsqwhHMZ7fx/4uH6Mk2wGXAO+pqusmuLapZjxj9XzgTd3v8w8Bv5vktIktb2IYRNtbAsxP8mtJnkDvRv+L19nmYmD4KcqjgCurdxfyxcAx3ROrvwbMBxZPUt2TbZPHKcn29H5x/XlVfWPSKm5jk8epql5YVUNVNQScAXygqqbuk5bjN55/e5cC+yaZ0wWvFwE3T1Ldk2084/QvwCHp2Rp4HnDLJNU92cYyTiO5FHhpkh2S7EDvqs2lm6nO1jZ5nLrtLwQ+U1Xnb8Yap4pNHquqem1V7d79Pn8bvTH7L0/dTwmtn5byUwC/BXyH3r0u7+7a/gp4Zbe8Jb2nmL9LL2g+tW/fd3f73Qq8vPW5TMVxAt5D7z615X2fX219PlNtnNbp4xSm+VPz4x0r4HX0HupaBXyw9blMxXECtunab6IX1N/e+lwaj9Nz6M1y3Q/8GLipb9/ju/H7LvB7rc9lKo5T92/u5+v8Ll/Q+nym4lit08dxTOGn5n2zkiRJkprw0rwkSZKaMIhKkiSpCYOoJEmSmjCISpIkqQmDqCRJkpowiErSFJPkvkk+3lCS35nMY0oSGEQlaUbr/iD/EL23IEnSpPJd85I0RSU5GHgf8O/AAuCLwErgrcBWwOFVdXuSs4GHgL2AnYE/raovJ9kS+N/0Xqv5SNd+VZLjgMPo/SH6rYE5wDOTLAfOoff2mn/o1gG8qar+uavnFOBHwN7AMuB1VVVJngN8pNvnZ8BvAA8ApwEHA7OBj1XV30/0OEkaXAZRSZra9gOeSe/d7XcAn6qqA5K8FXgz8CfddkP0XjX6NOCqJE8HTgSoqn2SPAO4LMke3fbPB/atqnu6gPm2qnoFQJI5wEuq6qEk84HP88t3xD+LXuD9Pr13xx+YZDFwHnB0VS3p3gf+IPAGYE1VPSfJbOAbSS6rqjs3wzhJGkAGUUma2pZU1Q8AktwOXNa1rwRe3LfdP1bVo8BtSe4AngEcBHwUoKpuSXIXMBxEv1pV94xwzMcDZyZZAKzt2wdgcVV9r6tnOb0AvAb4QVUt6Y710279S4F9kxzV7TsXmA8YRCUBBlFJmup+1rf8aN/3R3ns7/B139dcQDbQ7/0bWHcSvdsB9qP3LMFDI9Sztqsh6zk+Xfubq+rSDRxL0gzmw0qSND28OsnjkjwNeCpwK3AN8FqA7pL87l37uu4Ftu37PpfeDOejwOuBWaMc+xbgyd19oiTZtnsI6lLgj5I8friGJFtvoB9JM4wzopI0PdwKfI3ew0pv7O7v/DjwiSQr6T2sdFxV/Sz5LxOlK4BHktwInA18HLggyauBq9jw7ClV9XCSo4GPJtmK3v2hhwKfonfp/vr0DvofwOETcbKSpodUre9qiiRpUHRPzX+5qr7QuhZJ2hhempckSVITzohKkiSpCWdEJUmS1IRBVJIkSU0YRCVJktSEQVSSJElNGEQlSZLUhEFUkiRJTfx/nYNFb5ZBEVsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x = list(zip(train.columns, GBoost.feature_importances_))\n",
    "x.sort(key = lambda x: x[1])\n",
    "df = pd.DataFrame(x, columns = ['Var','Importance'])\n",
    "top = df[\"Importance\"][-10:]\n",
    "top_name = df['Var'][-10:]\n",
    "plt.figure(figsize = (10, 10))\n",
    "plt.barh(np.arange(len(top_name)), top)\n",
    "plt.yticks(np.arange(len(top_name))+ 0.25, np.array(top_name))\n",
    "_ = plt.xlabel('Importance')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Kernel Ridge was another model I tried out but ended up tossing due to low starting scores- I didn't want to bother spending the time on a model that wasn't performing relatively well to begin with"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "KRR = KernelRidge(alpha=0.6, kernel='polynomial', degree=2, coef0=2.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Kernel Ridge score: 0.1185 (0.0134)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "score = rmsle_cv(KRR)\n",
    "print(\"Kernel Ridge score: {:.4f} ({:.4f})\\n\".format(score.mean(), score.std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "LGBoost = GradientBoostingRegressor(n_estimators=1000, learning_rate=0.05,\n",
    "                                   max_depth=2, max_features='sqrt',\n",
    "                                    min_samples_leaf=15, min_samples_split=10,\n",
    "                                    loss='huber', random_state =5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LGBoost score: 0.1139 (0.0148)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "score = rmsle_cv(LGBoost)\n",
    "print(\"LGBoost score: {:.4f} ({:.4f})\\n\" .format(score.mean(), score.std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GradientBoostingRegressor(alpha=0.9, criterion='friedman_mse', init=None,\n",
       "             learning_rate=0.05, loss='huber', max_depth=2,\n",
       "             max_features='sqrt', max_leaf_nodes=None,\n",
       "             min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "             min_samples_leaf=15, min_samples_split=10,\n",
       "             min_weight_fraction_leaf=0.0, n_estimators=1000,\n",
       "             n_iter_no_change=None, presort='auto', random_state=5,\n",
       "             subsample=1.0, tol=0.0001, validation_fraction=0.1, verbose=0,\n",
       "             warm_start=False)"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "LGBoost.fit(X_train, Y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x = list(zip(train.columns, LGBoost.feature_importances_))\n",
    "x.sort(key = lambda x: x[1])\n",
    "df = pd.DataFrame(x, columns = ['Var','Importance'])\n",
    "top = df[\"Importance\"][-10:]\n",
    "top_name = df['Var'][-10:]\n",
    "plt.figure(figsize = (10, 10))\n",
    "plt.barh(np.arange(len(top_name)), top)\n",
    "plt.yticks(np.arange(len(top_name))+ 0.25, np.array(top_name))\n",
    "_ = plt.xlabel('Importance')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Random forest- this one also underperformed out of the gate and also took relatively long to run compared to my linear models so I ended up discarding it instead of trying to tune it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "rfr = rfr(random_state=0,\n",
    "        n_estimators=1000, max_depth=6,  max_features='sqrt')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rfr score: 0.1609 (0.0153)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "score = rmsle_cv(rfr)\n",
    "print(\"rfr score: {:.4f} ({:.4f})\\n\" .format(score.mean(), score.std()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Submitting the original models to kaggle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "     steps=[('robustscaler', RobustScaler(copy=True, quantile_range=(25.0, 75.0), with_centering=True,\n",
       "       with_scaling=True)), ('elasticnet', ElasticNet(alpha=0.0005, copy_X=True, fit_intercept=True, l1_ratio=0.9,\n",
       "      max_iter=1000, normalize=False, positive=False, precompute=False,\n",
       "      random_state=1, selection='cyclic', tol=0.0001, warm_start=False))])"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ENet.fit(train.values, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "final = ENet.predict(test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "results = np.exp(ENet.predict(test))\n",
    "# results = np.expm1(ENet.predict(test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "attempt = pd.DataFrame()\n",
    "attempt['Id'] = test_ID\n",
    "attempt['SalePrice'] = results\n",
    "attempt.to_csv('submit.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Ensembling by Averaging"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "the goal was to try and grab 2 random forest, 1 or 2 regression and 2 or 3 gbr. started with the models used in the kaggle jumpstart but liked my models better- discarded the random forest because it was underperforming compared to these other models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "class AveragingModels(BaseEstimator, RegressorMixin, TransformerMixin):\n",
    "    def __init__(self, models):\n",
    "        self.models = models\n",
    "    \n",
    "    #cloning models to fit the data\n",
    "    def fit(self, X, y):\n",
    "        self.models_ = [clone(x) for x in self.models]\n",
    "        \n",
    "        #training the cloned models\n",
    "        for model in self.models_:\n",
    "            model.fit(X, y)\n",
    "            \n",
    "        return self\n",
    "    \n",
    "    #predicting based on the cloned models\n",
    "    def predict(self, X):\n",
    "        predictions = np.column_stack([\n",
    "            model.predict(X) for model in self.models_])\n",
    "        #averaging the predictions\n",
    "        return np.mean(predictions, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "import random\n",
    "\n",
    "def feature_importances(clf, X, y):\n",
    "    score = np.mean(cross_val_score(clf, X,y,scoring='roc_auc'))\n",
    "    importances = {} \n",
    "    for i in range(X.shape[1]):\n",
    "        X_perm = X.copy()\n",
    "        X_perm[:,i] = random.sample(X[:,i].tolist(), X.shape[0])\n",
    "        perm_score = np.mean(cross_val_score(clf, X_perm , y, scoring='roc_auc'))\n",
    "        importances[i] = score - perm_score\n",
    "\n",
    "    return importances"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "averaged_models = AveragingModels(models = (ENet, GBoost, lasso, LGBoost))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Averaged base models score: 0.1085 (0.0142)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "score = rmsle_cv(averaged_models)\n",
    "print(\" Averaged base models score: {:.4f} ({:.4f})\\n\".format(score.mean(), score.std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AveragingModels(models=(Pipeline(memory=None,\n",
       "     steps=[('robustscaler', RobustScaler(copy=True, quantile_range=(25.0, 75.0), with_centering=True,\n",
       "       with_scaling=True)), ('elasticnet', ElasticNet(alpha=0.0005, copy_X=True, fit_intercept=True, l1_ratio=0.9,\n",
       "      max_iter=1000, normalize=False, positive=False...      subsample=1.0, tol=0.0001, validation_fraction=0.1, verbose=0,\n",
       "             warm_start=False)))"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "averaged_models.fit(train.values, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "finalstack = averaged_models.predict(test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "resultstack = np.exp(averaged_models.predict(test))\n",
    "# resultstack = np.expm1(averaged_models.predict(test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "attempt = pd.DataFrame()\n",
    "attempt['Id'] = test_ID\n",
    "attempt['SalePrice'] = resultstack\n",
    "attempt.to_csv('stacksubmit.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "AMMI",
   "language": "python",
   "name": "ammi"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}