ch15-code-listing.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Chapter 15: Machine learning"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Robert Johansson\n",
    "\n",
    "Source code listings for [Numerical Python - Scientific Computing and Data Science Applications with Numpy, SciPy and Matplotlib](https://www.apress.com/us/book/9781484242452) (ISBN 978-1-484242-45-2)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn import datasets\n",
    "from sklearn import model_selection\n",
    "from sklearn import linear_model\n",
    "from sklearn import metrics\n",
    "from sklearn import tree\n",
    "from sklearn import neighbors\n",
    "from sklearn import svm\n",
    "from sklearn import ensemble\n",
    "from sklearn import cluster"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Built in datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<function sklearn.datasets.base.load_boston(return_X_y=False)>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "datasets.load_boston"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<function sklearn.datasets.california_housing.fetch_california_housing(data_home=None, download_if_missing=True, return_X_y=False)>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "datasets.fetch_california_housing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<function sklearn.datasets.samples_generator.make_regression(n_samples=100, n_features=100, n_informative=10, n_targets=1, bias=0.0, effective_rank=None, tail_strength=0.5, noise=0.0, shuffle=True, coef=False, random_state=None)>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "datasets.make_regression"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "np.random.seed(123)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "X_all, y_all = datasets.make_regression(n_samples=50, n_features=50, n_informative=10) #, noise=2.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "X_train, X_test, y_train, y_test = model_selection.train_test_split(X_all, y_all, train_size=0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((25, 50), (25,))"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((25, 50), (25,))"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.shape, y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "model = linear_model.LinearRegression()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearRegression(copy_X=True, fit_intercept=True, n_jobs=None,\n",
       "         normalize=False)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def sse(resid):\n",
    "    return sum(resid**2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.469187171658096e-24"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "resid_train = y_train - model.predict(X_train)\n",
    "sse_train = sse(resid_train)\n",
    "sse_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.469187171658096e-24"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "resid_test = y_test - model.predict(X_test)\n",
    "sse_test = sse(resid_train)\n",
    "sse_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.score(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.31407400675201735"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def plot_residuals_and_coeff(resid_train, resid_test, coeff):\n",
    "    fig, axes = plt.subplots(1, 3, figsize=(12, 3))\n",
    "    axes[0].bar(np.arange(len(resid_train)), resid_train)\n",
    "    axes[0].set_xlabel(\"sample number\")\n",
    "    axes[0].set_ylabel(\"residual\")\n",
    "    axes[0].set_title(\"training data\")\n",
    "    axes[1].bar(np.arange(len(resid_test)), resid_test)\n",
    "    axes[1].set_xlabel(\"sample number\")\n",
    "    axes[1].set_ylabel(\"residual\")\n",
    "    axes[1].set_title(\"testing data\")\n",
    "    axes[2].bar(np.arange(len(coeff)), coeff)\n",
    "    axes[2].set_xlabel(\"coefficient number\")\n",
    "    axes[2].set_ylabel(\"coefficient\")\n",
    "    fig.tight_layout()\n",
    "    return fig, axes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x216 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plot_residuals_and_coeff(resid_train, resid_test, model.coef_)\n",
    "fig.savefig(\"ch15-regression-ols.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "model = linear_model.Ridge() #alpha=2.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ridge(alpha=1.0, copy_X=True, fit_intercept=True, max_iter=None,\n",
       "   normalize=False, random_state=None, solver='auto', tol=0.001)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "178.50695164950986"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "resid_train = y_train - model.predict(X_train)\n",
    "sse_train = sum(resid_train**2)\n",
    "sse_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "212737.00160105852"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "resid_test = y_test - model.predict(X_test)\n",
    "sse_test = sum(resid_test**2)\n",
    "sse_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.9994595515017335, 0.31670332736075457)"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.score(X_train, y_train), model.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x216 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plot_residuals_and_coeff(resid_train, resid_test, model.coef_)\n",
    "fig.savefig(\"ch15-regression-ridge.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "model = linear_model.Lasso(alpha=1.0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Lasso(alpha=1.0, copy_X=True, fit_intercept=True, max_iter=1000,\n",
       "   normalize=False, positive=False, precompute=False, random_state=None,\n",
       "   selection='cyclic', tol=0.0001, warm_start=False)"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "309.7497138953223"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "resid_train = y_train - model.predict(X_train)\n",
    "sse_train = sse(resid_train)\n",
    "sse_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1489.1176065002487"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "resid_test = y_test - model.predict(X_test)\n",
    "sse_test = sse(resid_test)\n",
    "sse_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x216 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plot_residuals_and_coeff(resid_train, resid_test, model.coef_)\n",
    "fig.savefig(\"ch15-regression-lasso.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "alphas = np.logspace(-4, 2, 100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/linear_model/coordinate_descent.py:492: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/linear_model/coordinate_descent.py:492: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/linear_model/coordinate_descent.py:492: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/linear_model/coordinate_descent.py:492: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/linear_model/coordinate_descent.py:492: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/linear_model/coordinate_descent.py:492: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/linear_model/coordinate_descent.py:492: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/linear_model/coordinate_descent.py:492: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/linear_model/coordinate_descent.py:492: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/linear_model/coordinate_descent.py:492: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/linear_model/coordinate_descent.py:492: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/linear_model/coordinate_descent.py:492: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/linear_model/coordinate_descent.py:492: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/linear_model/coordinate_descent.py:492: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/linear_model/coordinate_descent.py:492: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    }
   ],
   "source": [
    "coeffs = np.zeros((len(alphas), X_train.shape[1]))\n",
    "sse_train = np.zeros_like(alphas)\n",
    "sse_test = np.zeros_like(alphas)\n",
    "\n",
    "for n, alpha in enumerate(alphas):\n",
    "    model = linear_model.Lasso(alpha=alpha)\n",
    "    model.fit(X_train, y_train)\n",
    "    coeffs[n, :] = model.coef_\n",
    "    resid = y_train - model.predict(X_train)\n",
    "    sse_train[n] = sum(resid**2)\n",
    "    resid = y_test - model.predict(X_test)\n",
    "    sse_test[n] = sum(resid**2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(1, 2, figsize=(12, 4), sharex=True)\n",
    "\n",
    "for n in range(coeffs.shape[1]):\n",
    "    axes[0].plot(np.log10(alphas), coeffs[:, n], color='k', lw=0.5)\n",
    "\n",
    "axes[1].semilogy(np.log10(alphas), sse_train, label=\"train\")\n",
    "axes[1].semilogy(np.log10(alphas), sse_test, label=\"test\")\n",
    "axes[1].legend(loc=0)\n",
    "\n",
    "axes[0].set_xlabel(r\"${\\log_{10}}\\alpha$\", fontsize=18)\n",
    "axes[0].set_ylabel(r\"coefficients\", fontsize=18)\n",
    "axes[1].set_xlabel(r\"${\\log_{10}}\\alpha$\", fontsize=18)\n",
    "axes[1].set_ylabel(r\"sse\", fontsize=18)\n",
    "fig.tight_layout()\n",
    "fig.savefig(\"ch15-regression-lasso-vs-alpha.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "model = linear_model.LassoCV()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2053: FutureWarning: You should specify a value for 'cv' instead of relying on the default value. The default value will change from 3 to 5 in version 0.22.\n",
      "  warnings.warn(CV_WARNING, FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "LassoCV(alphas=None, copy_X=True, cv='warn', eps=0.001, fit_intercept=True,\n",
       "    max_iter=1000, n_alphas=100, n_jobs=None, normalize=False,\n",
       "    positive=False, precompute='auto', random_state=None,\n",
       "    selection='cyclic', tol=0.0001, verbose=False)"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(X_all, y_all)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.06559238747534718"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.alpha_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.5450589323146204"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "resid_train = y_train - model.predict(X_train)\n",
    "sse_train = sse(resid_train)\n",
    "sse_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.5321417406216717"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "resid_test = y_test - model.predict(X_test)\n",
    "sse_test = sse(resid_test)\n",
    "sse_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.9999953221722068, 0.9999950788657098)"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.score(X_train, y_train), model.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x216 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plot_residuals_and_coeff(resid_train, resid_test, model.coef_)\n",
    "fig.savefig(\"ch15-regression-lasso-cv.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "model = linear_model.ElasticNetCV()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2053: FutureWarning: You should specify a value for 'cv' instead of relying on the default value. The default value will change from 3 to 5 in version 0.22.\n",
      "  warnings.warn(CV_WARNING, FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "ElasticNetCV(alphas=None, copy_X=True, cv='warn', eps=0.001,\n",
       "       fit_intercept=True, l1_ratio=0.5, max_iter=1000, n_alphas=100,\n",
       "       n_jobs=None, normalize=False, positive=False, precompute='auto',\n",
       "       random_state=None, selection='cyclic', tol=0.0001, verbose=0)"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(X_all, y_all)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.13118477495069433"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.alpha_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.l1_ratio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2183.8391729391196"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "resid_train = y_train - model.predict(X_train)\n",
    "sse_train = sum(resid_train**2)\n",
    "sse_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2650.050446338252"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "resid_test = y_test - model.predict(X_test)\n",
    "sse_test = sum(resid_test**2)\n",
    "sse_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.9933881981034111, 0.9914882195448783)"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.score(X_train, y_train), model.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA1gAAADQCAYAAAAalMCAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzt3XucbHV55/vPl5s3MIhuCALbjQ5e0COoHaJiHERUVBQ14CXqoCFDjHjEURNBZ4Lx6AzkeImjRtxBBBEVvB0QjYpEZTSKbBDkppEQkB22bFAZ8DLohuf8UaulaLr37t1dVWtV1ef9evWral1qradW93q6nvX7rV+lqpAkSZIkLd8WbQcgSZIkSZPCAkuSJEmSBsQCS5IkSZIGxAJLkiRJkgbEAkuSJEmSBsQCS5IkSZIGxAJLS5LkhCT/bdDrLleSa5IcMIp9SeoO84wkqSu2ajsAjV6Sa4A/q6qvLnUbVfWqYaw7SkkK2KOqrmo7FmkaDCL3NNt5RbOdJ83OM89IkrrCFizdTRILb0mSJGkJLLCmTJJTgZXA55P8IslfJVmVpJIcnuTHwD81634qyU+S/O8k5yV5ZN92Tk7y9ub5fknWJnlDkvVJ1iV55RLXvX+Szye5JckFSd6e5JsbeT8vT3Jtkp8mecucZfsk+XaSm5v9vD/JNs2y85rVLmmOw4uS3C/J2UluTPLz5vmuyz3mkubPPc38xyf55+Y8vSTJfn2veUWSq5PcmuTfkrw0ySOAE4AnNNu5uVnXPCNJ6gQLrClTVS8Hfgw8p6q2raq/7Vv8H4FHAM9opv8R2APYEbgIOG0jm/594PeAXYDDgQ8kud8S1v0A8MtmncOan3kl2RP4IPBy4IHA/YH+Dyq3A/8FeADwBOCpwKub4/DkZp29muNwOr3z4SPAg+h9EPw18P6NvGdJizRf7kmyC/AF4O3ADsAbgc8kWZHkPsD/BJ5ZVdsBTwQurqorgVcB3262s/0CuzTPSJJaYYGlfm+tql9W1a8Bquqkqrq1qm4D3grsleT3Fnjtb4G3VdVvq+qLwC+Ah23Oukm2BP4YOLaqflVVVwCnbCTeQ4Czq+q8Jsb/Btwxu7CqLqyq71TVhqq6BvgQvSJyXlX106r6TLPvW4F3bGx9Scv2MuCLVfXFqrqjqs4B1gDPapbfATwqyb2qal1VXb4Z2zbPSJJaYYGlftfNPkmyZZLjkvxrkluAa5pFD1jgtT+tqg19078Ctt3MdVfQG3jlur5l/c/nemD/8qr6JfDTvvfw0Kb7zU+a9/DfNxI/Se6d5ENNV6BbgPOA7ZsPZJIG70HAoU33upub7n5PAnZuzucX0WutWpfkC0kevhnbNs9IklphgTWdahHz/wQ4GDiAXjebVc38DC8sbgQ2cNfuN7ttZP11/cuT3Jte951ZHwR+QG8Er/sCb2bj8b+BXqvbHzbrz3bvGeZ7lqbJ3NxzHXBqVW3f93OfqjoOoKq+XFVPA3amdy7/wwLb2RzmGUnSUFlgTacbgAdvYp3tgNvoXam9N72rskNVVbcDnwXe2lzlfTjwnzbykk8DByV5UnNT+du469/0dsAtwC+abf3FnNfPPQ7b0bsf4uYkOwDHLusNSZpr7jn3MeA5SZ7RtJrfsxmgYtckOyV5bnMv1m30uvjd3redXWcHk9gc5hlJ0rBZYE2n/wH816ZLzhsXWOejwLXAvwNXAN8ZUWyvoddi9hPgVOAT9D5c3U1zP8aRwMfpXWX+ObC2b5U30muJu5Xele/T52zircApzXF4IfB3wL2Am+i93y8N5B1JmnWX3FNV19FrKX8zvZal64C/pPe/aQt6rT3XAz+jd5/Sq5vt/BNwOfCTJDctIQ7zjCRpaFK1nJ4W0nAlOR74/apacJQvSVoO84wkaZBswVKnJHl4kkenZx96wyt/ru24JE0O84wkaZi2ajsAaY7t6HXXeSCwHngXcGarEUmaNOYZSdLQ2EVQkiRJkgbELoKSJEmSNCAT0UXwAQ94QK1atartMCQBF1544U1VtaLtOIbJnCN1x6TnHPON1B2LzTcTUWCtWrWKNWvWtB2GJCDJtW3HMGzmHKk7Jj3nmG+k7lhsvpmIAkuSJKlLklxD7/vRbgc2VNVM8+XSpwOrgGuAF1bVz9uKUdJweA+WJEnScDylqvauqplm+mjg3KraAzi3mZY0YVorsJLsluRrSa5McnmSo5r5OyQ5J8mPmsf7tRWjJEnSAB0MnNI8PwV4XouxSBqSNluwNgBvqKpHAI8HjkyyJ17dkSRJ46+AryS5MMkRzbydqmodQPO443wvTHJEkjVJ1tx4440jClfSoLRWYFXVuqq6qHl+K3AlsAte3ZEkSeNv36p6LPBMeheRn7zYF1bV6qqaqaqZFSsmdoBEaWJ1YpCLJKuAxwDnM+fqTpJ5r+5Ii7Xq6C8suOya4549wkgkLcdC57Lnsbqoqq5vHtcn+RywD3BDkp2bzzc7A+tbDXKO/nNsWOfVKPYhta31QS6SbAt8BnhdVd2yGa+z+VySJHVOkvsk2W72OfB04DLgLOCwZrXDgDPbiVDSMLXagpVka3rF1WlV9dlm9qKu7lTVamA1wMzMTI0kYEmSpE3bCfhcEuh91vp4VX0pyQXAGUkOB34MHNpijJKGpLUCK72s82Hgyqp6d9+i2as7x+HVHUmSNGaq6mpgr3nm/xR46ugjkjRKbbZg7Qu8HLg0ycXNvDfTK6y8uiNJkqSJ5f1ok6u1AquqvglkgcVe3dFYcSANSZIkQQcGuZAkSZKkSWGBJUmSJEkDYoElaSok2S3J15JcmeTyJEc183dIck6SHzWP92s7VkmSNL4ssCRNiw3AG6rqEcDjgSOT7AkcDZxbVXsA5zbTkiRJS2KBJWkqVNW6qrqoeX4rcCWwC3AwcEqz2inA89qJUJIkTQILLElTJ8kq4DHA+cBOVbUOekUYsOMCrzkiyZoka2688cZRhSpJksaMBZakqZJkW+AzwOuq6pbFvq6qVlfVTFXNrFixYngBSpKksWaBJWlqJNmaXnF1WlV9tpl9Q5Kdm+U7A+vbik+SJI0/CyxJUyFJgA8DV1bVu/sWnQUc1jw/DDhz1LFJkqTJsVXbAUjSiOwLvBy4NMnFzbw3A8cBZyQ5HPgxcGhL8UmSpAlggSVpKlTVN4EssPipo4xFkiRNLrsISpIkSdKAWGBJkiQNQZItk3wvydnN9O5Jzk/yoySnJ9mm7RglDZ4FliRJ0nAcRe9LzWcdD7ynqvYAfg4c3kpUkobKAkuSJGnAkuwKPBs4sZkOsD/w6WaVU4DntROdpGGywJIkSRq8vwP+Crijmb4/cHNVbWim1wK7zPfCJEckWZNkzY033jj8SCUNlAWWJEnSACU5CFhfVRf2z55n1Zrv9VW1uqpmqmpmxYoVQ4lR0vA4TLskSdJg7Qs8N8mzgHsC96XXorV9kq2aVqxdgetbjFHSkFhgSVKHrTr6C/POv+a4Z484EkmLVVXHAMcAJNkPeGNVvTTJp4BDgE8ChwFnthakpKGxi6AkSdJovAl4fZKr6N2T9eGW45E0BLZgSZIkDUlVfR34evP8amCfNuORNHy2YEmSJEnSgFhgSZIkSdKA2EVQkiQN3EIDtICDtEiabLZgSZIkSdKAWGBJkiRJ0oBYYEmSJEnSgFhgSZIkSdKAtDrIRZKTgIOA9VX1qGbeDsDpwCrgGuCFVfXztmKUJG2cgxlIknSntluwTgYOnDPvaODcqtoDOLeZliRJkqTOa7XAqqrzgJ/NmX0wcErz/BTgeSMNSpIkSZKWqIvfg7VTVa0DqKp1SXacb6UkRwBHAKxcuXKE4S2f3WkkSZKkydR2F8Elq6rVVTVTVTMrVqxoOxxJkiRJ6mSBdUOSnQGax/UtxyNpAiQ5Kcn6JJf1zdshyTlJftQ83q/NGCVJ0vjrYoF1FnBY8/ww4MwWY5E0OU7GQXUkSdKQtVpgJfkE8G3gYUnWJjkcOA54WpIfAU9rpiVpWRxUR9KoJLlnku8muSTJ5Un+ppm/e5Lzm1bz05Ns03askgav1UEuquolCyx66kgDkTStFjWoDoz3wDqaLA6UNBZuA/avql8k2Rr4ZpJ/BF4PvKeqPpnkBOBw4INtBipp8LrYRVCSOseBdaTJk2TfxczbXNXzi2Zy6+angP2BTzfzbTWXJpQFlqRp5qA60nR73yLnbbYkWya5mF5eOQf4V+DmqtrQrLIW2GUQ+5LULV38HixJGpXZQXWOw0F1pKmR5AnAE4EVSV7ft+i+wJaD2EdV3Q7snWR74HPAI+ZbbYH47JIsjTELLHWa9xpoUJpBdfYDHpBkLXAsvcLqjGaAnR8Dh7YXoaQR2gbYlt7noO365t8CHDLIHVXVzUm+Djwe2D7JVk0r1q7A9Qu8ZjWwGmBmZmbeIkxSd1lgSZoKDqojaVZVfQP4RpKTq+raQW8/yQrgt01xdS/gAOB44Gv0CrhPYqu5NLEssCRJ0rS6R5LVwCr6PhNV1f7L3O7OwClJtqR3v/sZVXV2kiuATyZ5O/A94MPL3I+kDpq6AmuhLmd2N5Mkaep8CjgBOBG4fVAbrarvA4+ZZ/7VwD6D2o+kbtpogZXksRtbXlUXDTYcSdPKfCOpBRuqyu+hkjRQm2rBetdGls1+n4MkDYL5RtKofT7Jq+mN8nfb7Myq+ll7IUkadxstsKrqKaMKRNJ0m8R84yiYUucd1jz+Zd+8Ah7cQiySJsSi78FK8ihgT+Ces/Oq6qPDCErSdDPfSBqFqtq97RgkTZ5FFVhJjqX3/TF7Al8Engl8E/ADjzbKQUXGSxdaXMw3kkYlyb2B1wMrq+qIJHsAD6uqs1sOTdIYW2wL1iHAXsD3quqVSXaiN+KOJA2a+Uad1YWLEBqojwAXAk9sptfSG1nQAkvSki22wPp1Vd2RZEOS+wLrsX+ypOEw32hsWYCNnYdU1YuSvASgqn6dJG0HJWm8LbbAWpNke+Af6F3p+QXw3aFF1SL/OUqtm5p80xV25dUU+02Se9Eb2IIkD6FvNEFJk2nu/71B/79bVIFVVa9unp6Q5EvAfZsv0ZOkgTLfSBqhY4EvAbslOQ3YF3hFqxFJGnuLHeTiyfPNq6rzBh+SpGk2TfnGFnN12TS0bFbVOUkuAh4PBDiqqm5qOSxJY26xXQT7vx/insA+9Lru+MWfkgbNfCNpqJI8vKp+kOSxzax1zePKJCur6qK2YpM0/hbbRfA5/dNJdgP+digRSZpq5htJI/B64AjgXfMsK7ygI2kZFv1Fw3OsBR41yEAkaQHmG0kDVVVHNI9PaTsWSZNnsfdgvY9mhB1gC2Bv4JJhBSVpeplvusd7xTSpkhwJnFZVNzfT9wNeUlV/325kksbZoodp73u+AfhEVX1rCPFIkvlG0qj856r6wOxEVf08yX8GLLAkLdli78E6ZdiBSNNuGkbsWgzzjaQR2iJJqmr2e7C2BLZZ7kabe0c/Cvw+cAewuqrem2QH4HRgFXAN8MKq+vly9yepWzZaYCW5lDu76txNVT164BFJmkrmG0kt+DJwRpIT6OWfV9H7Xqzl2gC8oaouSrIdcGGSc+h9x9a5VXVckqOBo4E3DWB/kjpkUy1YBzWPRzaPpzaPLwV+NZSIJE0r842mgq3VnfIm4M+Bv6D3PVhfAU5c7karah3N0O9VdWuSK4FdgIOB/ZrVTgG+jgWWNHE2WmBV1bUASfatqn37Fh2d5FvA24YZnKTpYb6Rpk/bxWZV3QF8sPkZiiSrgMcA5wM7NcUXVbUuyY4LvOYIesPIs3LlymGFpgHq/1v2Yom2WOR690nypNmJJE8E7jOckCRNOfONpKFKckbzeGmS78/9GeB+tgU+A7yuqm5Z7OuqanVVzVTVzIoVKwYVjqQRWewogocDJyX5vWb6ZuBPhxNST5IDgfcCWwInVtVxw9yfpM4w30gattc1jwdtdK1lSLI1veLqtKr6bDP7hiQ7N61XOwPrh7V/Se1Z7CiCFwJ7JbkvkKr638MMqhnF5wPA0+h9yegFSc6qqiuGuV9J7TPfSBqBs4HHAm+vqpcPeuNJAnwYuLKq3t236CzgMOC45vHMQe9bUvs2NYrgy6rqY0leP2c+AHOSxiDtA1xVVVc3+/skvRtD/cDTIX75qAbJfCNphLZJchjwxCQvmLuwr8VpqfYFXg5cmuTiZt6b6RVWZyQ5HPgxcOgy9yOpgzbVgjV738N2ww5kjl2A6/qm1wJ/2L/CUm8A3dQH/00tX0xRsambdkexj01tYzn7mF1nMUXUct/rYvbRheM9iH104e+i5cK4s/lG0sR5Fb0RSrcHnjNnWQHLKrCq6pv0RiWcz1OXs21J3bepUQQ/1Dz+zWjC+Z35ktJdvh+nqlYDqwFmZmYW/O4caZqMc8thl/MNLO2iznIv6CxmnUFcTFlunMu5EDL7+lFchBjFxZSuX3ia3cYojvcgjsUQ7VxVf5Hke83nCUkamEXdg5Xkb4G3A7+m9wV8e9EbEedjQ4prLbBb3/SuwPVD2pekDulqvvGizvgb5wsQGrhjgE/Ra8mywJI0UIsdpv3pzfCiB9H7MPJQ4C+HFhVcAOyRZPck2wAvpndjqKTJZ76RNGw/TfI1YPckZ839aTs4SeNtscO0b908Pgv4RFX9bPbG82Goqg1JXgN8md6wySdV1eVD26GkLjHfSBq2Z9MbRfBU4F0txyJpwiy2wPp8kh/Q67Lz6iQrgP8zvLCgqr4IfHGY+5DUSeYbaQyMc5fLqvoN8J0kT6yqG5Pcp6p+2XZckibDYr8H6+gkxwO3VNXtSX5FbxhjSRoo881kGucP45po/yHJN4BtgZVJ9gL+vKpe3XJcksbYYge5uDdwJLCS3ihaDwQeRu+L+tQCP6xoUplvJI3Q3wHPoLnvsqouSfLkdkOSNO4W20XwI8CFwBOb6bX0Rt/xA4+kQTPfaKp5AW20quq6Ofd53t5WLJImw2ILrIdU1YuSvASgqn6dYd51LulupuhDl/lG0qhcl+SJQDWjiL4WuLLlmCSNucUWWL9Jci+aL99M8hDgtqFFJWmamW+m0BRdQFC3vAp4L7AL8O/0RhM9stWIJI29TRZYzZXjE+h94eduSU4D9gVeMdzQpMXxg9nkMN9IGqWqugl4adtxSJosmyywqqqSHAU8HXg8EOCoJilJ0sCYbySNUpJdgffRu5BTwDfp5Zy1rQYmaawttovgd4AHV9UXhhnMpLBFRVoW842kUfkI8HHg0Gb6Zc28p7UWkaSxt9gC6ynAnye5FvglvavKVVWPHlpkkqaV+UbSqKyoqo/0TZ+c5HWtRaOhWHX0ndfrvAiuUVhsgfXMoUYhSXcy30hD5ofM37kpycuATzTTLwF+utyNJjkJOAhYX1WPaubtAJwOrAKuAV5YVT9f7r40uSwMx9cWi1mpqq6d72fYwUmaPuYbSSP0p8ALgZ8A64BDgFcOYLsnAwfOmXc0cG5V7QGc20xLmkCLKrAkSZIm0P8DHFZVK6pqR3oF11uXu9GqOg/42ZzZBwOnNM9PAZ633P1I6qbFdhGUpppN85I0kR7d302vqn6W5DFD2tdOVbWu2c+6JDsOaT+SWmaBJUmSptUWSe43W2Q190m1/tkoyRHAEQArV65sORpNEu/rGo3Wk4gkSVJL3gX8c5JP0/serBcC7xjSvm5IsnPTerUzsH6hFatqNbAaYGZmpoYUj6Qh8R4sSZI0larqo8AfAzcANwIvqKpTh7S7s4DDmueHAWcOaT+SWmYLliRJmlpVdQVwxSC3meQTwH7AA5KsBY4FjgPOSHI48GPu/HJjSRPGAkuSNBG8n0BdUVUvWWDRU0caiKRWWGBJkiRJY6Z/wArwIlOXWGBp6pmQJEmaXhYqk2/u73jYLLAkSZIkdda4DS9vgSVJkiRpbHS94LLAkkagiye/JEmSBs8CS9LES3Io8FbgEcA+VbWmb9kxwOHA7cBrq+rLrQQpDZAXdSSpPRZYkqbBZcALgA/1z0yyJ/Bi4JHAA4GvJnloVd0++hAlqVsc/GG0ut7tTYtngSVp4lXVlQBJ5i46GPhkVd0G/FuSq4B9gG+PNkJJ0qiMekQ5TR8LLEnTbBfgO33Ta5t5d5PkCOAIgJUrVw4/MkmaAsNoJbMlSG3boo2dJjk0yeVJ7kgyM2fZMUmuSvLDJM9oIz5J4yfJV5NcNs/PwRt72Tzzar4Vq2p1Vc1U1cyKFSsGE7QkaSysOvoLv/uRNqWtFizvh5A0UFV1wBJethbYrW96V+D6wUQkSdJ4a+M+vEm496+VAsv7ISR1xFnAx5O8m95FnT2A77YbkiRJk2tuF85JbBXs2j1Y3g8xIl25GtCVODTZkjwfeB+wAvhCkour6hlVdXmSM4ArgA3AkbaYS1J7JqH1Yli8t2x8DK3ASvJV4PfnWfSWqjpzoZfNM2/B+yGA1QAzMzPzriNJAFX1OeBzCyx7B/CO0UYkSRoUCw91zdAKLO+HaJcJRpIkSRq9rnUR9H4IaYgsvCWpXUkOBN4LbAmcWFXHtRzSsth6NFkm5ffZ9vtopcDyfghJkjRtkmwJfAB4Gr1eOxckOauqrhjG/tr+kClNq7ZGERzb+yFMUJIkaYn2Aa6qqqsBknyS3gjKQymwJLUjVeM/PsTMzEytWbOm7TAGZmPDVVrgqeuSXFhVM5tec3x1KedsKl+YTzbPQsfLY9Vd45RzkhwCHFhVf9ZMvxz4w6p6zZz1+kdKfty11167yW0vZvS9jQ2PvZi/8VHsYyn7HIZNxb05y+czjPcx37EadCvmYvYx6L+BzY1rmH8ji803WwwtAkmSJPVb1GjJVbW6qmaqambFihUjCEvSIFlgSZIkjYajJUtToGujCEqSJE2qC4A9kuwO/DvwYuBP2g1JwzC3m9qmugxqslhgSZIkjUBVbUjyGuDL9IZpP6mqLh/V/ud+6N/ce1W8F7Hb/P10hwWWJEnSiFTVF4Evth2HpOGxwJIkSdJA2Ioy2fz9Lo4FliRJkqSx1bXCzwJLkiRJ6tO1D+waLxZYkiRJ0hBZsE0XCyxJGmP+05YE5gJ11zT+bfpFw5IkSZI0IBZYkiRJkjQgdhGUJEmSNBDT2CVwLgssSZIkaTNYRGhj7CIoSZIkSQNigSVJkiRJA2IXQUmSJA3FcrvS2RVveTx+7bDAkjTxkvy/wHOA3wD/Cryyqm5ulh0DHA7cDry2qr7cWqBD4D9XSZJGywKrg/xAJA3cOcAxVbUhyfHAMcCbkuwJvBh4JPBA4KtJHlpVt7cYq1pk/pWkzWPevDvvwZI08arqK1W1oZn8DrBr8/xg4JNVdVtV/RtwFbBPGzFKmgxJDk1yeZI7kszMWXZMkquS/DDJM9qKUdJwWWBJmjZ/Cvxj83wX4Lq+ZWubeXeT5Igka5KsufHGG4ccoqQxdhnwAuC8/plzWswPBP4+yZajD0/SsFlgSZoISb6a5LJ5fg7uW+ctwAbgtNlZ82yq5tt+Va2uqpmqmlmxYsXg34CkiVBVV1bVD+dZZIu5NCW8B0vSRKiqAza2PMlhwEHAU6tqtohaC+zWt9quwPXDiVDSlNuFXhflWRttMQeOAFi5cuXwI5M0ULZgSZp4SQ4E3gQ8t6p+1bfoLODFSe6RZHdgD+C7bcQoaXwspsV8vpfNM88Wc2kC2YIlaRq8H7gHcE4SgO9U1auq6vIkZwBX0Os6eKQjCEralE21mC/AFnNpSuTOnjLjK8mNwLVLeOkDgJsGHM4wGOfgjEOMMN5xPqiqJvqS6xJzzjj/TrvIOAdrHOJcKMZO5pwkXwfeWFVrmulHAh+nd9/VA4FzgT02dVFnCflmHH6XMB5xjkOMYJyDtKkYF5VvJqLAWqoka6pqZtNrtss4B2ccYgTjnETjcqyMc7CMc3DGIUaAJM8H3gesAG4GLq6qZzTL3kJvJNMNwOuq6h8X3NDS9z8ux6nzcY5DjGCcgzSoGO0iKEmSNCBV9TngcwssewfwjtFGJGnUHORCkiRJkgZk2gus1W0HsEjGOTjjECMY5yQal2NlnINlnIMzDjF2wbgcp3GIcxxiBOMcpIHEONX3YEmSJEnSIE17C5YkSZIkDYwFliRJkiQNyFQWWEkOTPLDJFclObrteBaS5Joklya5OMmatuOZleSkJOuTXNY3b4ck5yT5UfN4vzZjbGKaL863Jvn35phenORZbcbYxLRbkq8luTLJ5UmOauZ35phuJMbOHc8uMucszzjkHPPNyOLs3DHtki7mmnE4f5uYOn9uJLlnku8muaSJ8W+a+bsnOb+J8fQk27QVY78kWyb5XpKzm+nOxTnf/71B/M6nrsBKsiXwAeCZwJ7AS5Ls2W5UG/WUqtq7Y98bcDJw4Jx5RwPnVtUe9L48sQuJ/WTuHifAe5pjundVfXHEMc1nA/CGqnoE8HjgyOZvskvHdKEYoXvHs1PMOQNxMt3POSdjvhkkc85m6nCuOZnun78wHufGbcD+VbUXsDdwYJLHA8fTOy/2AH4OHN5ijP2OAq7sm+5qnHP/7y37dz51BRa9b1C/qqqurqrfAJ8EDm45prFSVecBP5sz+2DglOb5KcDzRhrUPBaIs3Oqal1VXdQ8v5VeMtqFDh3TjcSoTTPnLNM45BzzzWCZc5akk7lmHM5fGI9zo3p+0Uxu3fwUsD/w6WZ+J45nkl2BZwMnNtOhg3EuYNm/82kssHYBruubXkt3k3YBX0lyYZIj2g5mE3aqqnXQS1LAji3HszGvSfL9pttC690S+iVZBTwGOJ+OHtM5MUKHj2dHmHOGo5Pnxzw6e36MQ74Bc85mGKdc09m/N+j2udF0u7sYWA+cA/wrcHNVbWhW6crv/e+AvwLuaKbvTzfjnO//3rJ/59NYYGWeeV0dq37fqnosveb+I5M8ue2AJsAHgYfQa1pfB7yr3XDulGRb4DPA66rqlrbjmc88MXb2eHaIOWd6dfb8GId8A+aczTROuaazun5uVNXtVbU3sCu9VstHzLfaaKO6qyQHAeur6sL+2fOs2oW/z6H835vGAmstsFvf9K7A9S3FslFVdX3zuB74HL0TqatuSLIzQPO4vuV45lVVNzTJ6Q7gH+jIMU2yNb2EflpVfbaZ3alRRgZiAAAH2UlEQVRjOl+MXT2eHWPOGY5OnR/z6er5MQ75ponDnLN5xibX0MG/NxifcwOgqm4Gvk7vfrHtk2zVLOrC731f4LlJrqHXVXV/ei1aXYtzof97y/6dT2OBdQGwRzOSyTbAi4GzWo7pbpLcJ8l2s8+BpwOXbfxVrToLOKx5fhhwZouxLGj2hGk8nw4c06Zf8oeBK6vq3X2LOnNMF4qxi8ezg8w5w9GZ82MhXTw/xiHfgDlnicYi1zQ69fcG43FuJFmRZPvm+b2AA+jdK/Y14JBmtdaPZ1UdU1W7VtUqen+H/1RVL6VjcW7k/97yf+dVNXU/wLOAf6HXb/UtbcezQIwPBi5pfi7vUpzAJ+h1zfgtvStmh9PrW3su8KPmcYeOxnkqcCnw/eYE2rkDcT6JXjP594GLm59ndemYbiTGzh3PLv6Yc5YdW+dzjvlmZHF27ph26aeLuWYczt8mzs6fG8Cjge81MV4G/HUz/8HAd4GrgE8B92j7ePbFvB9wdhfjXOj/3iB+52k2JEmSJElapmnsIihJkiRJQ2GBJUmSJEkDYoElSZIkSQNigSVJkiRJA2KBJUmSJEkDYoGloUry9SQzHYjjFUne33YckobLnCNpU5LcI8lXk1yc5EVJ/ijJ5c30Lkk+vYnXn5hkzyXue78kT1xa5EuTZFUSvzNuhLba9CqSkmxZVbe3HYek6WDOkYbqMcDWVbU3QJITgHdW1Uea5Ycs+Eqgqv5sGfveD/gF8M/L2MZIJdmqqja0Hcc4sQVryjTfWv2FJJckuSzJi5r5f53kgmbe6uYbzWevBr8nyXlJrkzyB0k+m+RHSd7erLMqyQ+SnJLk+0k+neTe8+z76Um+neSiJJ9Ksu0863w9yfFJvpvkX5L8UTP/LleDk5ydZL/m+S+a11zYXJHap9nO1Ume27f53ZJ8KckPkxzbt62XNfu7OMmHkmzZt923JTkfeMIADr80dcw55hxpkJL8p+a8vyTJqc28ByU5t5l/bpKVzfwVST7T5JoLkuybZEfgY8DezTn458ALgb9Oclr6WnuSbJnknUkubbb9fzfzf9dSvlCeSXJNkr9p5l+a5OFJVgGvAv5Ls+8/mvPe3prkpL588tpm/l1aoJK8Mclb+2LZaM5sbDVfzkzyuCTfaPLZl5Ps3Lfd/57kG8BRA/wVTgULrOlzIHB9Ve1VVY8CvtTMf39V/UEz717AQX2v+U1VPRk4ATgTOBJ4FPCKJPdv1nkYsLqqHg3cAry6f6dJHgD8V+CAqnossAZ4/QIxblVV+wCvA45dYJ1+9wG+XlWPA24F3g48DXg+8La+9fYBXgrsDRyaZCbJI4AXAfs2V7Jub9aZ3e5lVfWHVfXNRcQh6e7MOeYcaSCSPBJ4C7B/Ve3FnR/83w98tMkHpwH/s5n/XuA9VfUHwB8DJ1bVeuDPgP9VVXtX1YeAs4C/rKqXcldHALsDj+nbdn88m8ozNzXzPwi8saquoZfX3tPs+3/N8zYfDjyDXv44NsnWizg0S8qZzbbfBxzS5LOTgHf0bXf7qvqPVfWuRcSgPnYRnD6XAu9Mcjxwdt/J/ZQkfwXcG9gBuBz4fLPsrL7XXl5V6wCSXA3sBtwMXFdV32rW+xjwWuCdfft9PLAn8K30LlRvA3x7gRg/2zxeCKxaxHv6DXd+aLsUuK2qfpvk0jmvP6eqftrE/lngScAG4HHABU1c9wLWN+vfDnxmEfuXtDBzDuYcaUD2Bz5dVTcBVNXPmvlPAF7QPD8V+Nvm+QHAns25BnDfJNttxv4OAE6Y7R7Xt79Zm8oz/bnlBSzOF6rqNuC2JOuBnRbxmqXmzC/RK8TOaeLfEljXt93TFxmz5rDAmjJV9S9JHgc8C/gfSb5CLxH9PTBTVdc1zc737HvZbc3jHX3PZ6dn/4Zq7q7mTIfeh42XLCLM2X3c3rf9Ddy1xbU/vt9W1ez+fhdjVd2RpP9vfL4YA5xSVcfME8f/8R4IaXnMOXeL0ZwjLV24+3k1n9l1tgCeUFW/vstG7iy4lru/TeWZ+XLLpvTnvNnXbSwf9b9mc3Nm6BVkC3VJ/uUiY9YcdhGcMkkeCPyqqj5G72rvY7nzRL2p6Tu80Zs7F7AyyewJ+hJgbveW7wD7JvkPTRz3TvLQzdj+NfT6S2+RZDd6Teeb62lJdkhyL+B5wLeAc4FDmj7ZNMsftIRtS5qHOcecIw3QucALZ7u9Jdmhmf/PwIub5y/lznzwFeA1sy9Osvdm7u8rwKtmL5z07W/WUvLMrcDmtKIB3ADsmOT+Se7BXbtUL9Z8OfOHwIrZ+Um2brphapkssKbP/wV8N8nF9Poxv72qbgb+gV7T8v8HXLCE7V4JHJbk+/S6+3ywf2FV3Qi8AvhEs8536PUzXqxvAf/WxPhO4KIlxPhNel0HLgY+U1VrquoKev2nv9LEdQ6w8xK2LWl+5hxzjjQQVXU5vXuEvpHkEuDdzaLXAq9szqmXc+e9Wa8FZpqBHa6gN8DE5jgR+DHw/WZ/fzInnqXkmc8Dz888g1wspKp+S+/+zvOBs4EfbM6baNwtZ1bVb+hd4Dq+eX8XAyMdQn5S5c5eDtLSpDcqztnNzeqSNFTmHElSl9mCJUmSJEkDYguWJEmSJA2ILViSJEmSNCAWWJIkSZI0IBZYkiRJkjQgFliSJEmSNCAWWJIkSZI0IP8/T5xxj6k93acAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 864x216 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plot_residuals_and_coeff(resid_train, resid_test, model.coef_)\n",
    "fig.savefig(\"ch15-regression-elastic-net-cv.pdf\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "iris = datasets.load_iris()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "sklearn.utils.Bunch"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type(iris)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['setosa', 'versicolor', 'virginica'], dtype='<U10')"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "iris.target_names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['sepal length (cm)',\n",
       " 'sepal width (cm)',\n",
       " 'petal length (cm)',\n",
       " 'petal width (cm)']"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "iris.feature_names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(150, 4)"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "iris.data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(150,)"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "iris.target.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# print(iris['DESCR'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "X_train, X_test, y_train, y_test = model_selection.train_test_split(iris.data, iris.target, train_size=0.7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "classifier = linear_model.LogisticRegression()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/linear_model/logistic.py:433: FutureWarning: Default solver will be changed to 'lbfgs' in 0.22. Specify a solver to silence this warning.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/linear_model/logistic.py:460: FutureWarning: Default multi_class will be changed to 'auto' in 0.22. Specify the multi_class option to silence this warning.\n",
      "  \"this warning.\", FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
       "          intercept_scaling=1, max_iter=100, multi_class='warn',\n",
       "          n_jobs=None, penalty='l2', random_state=None, solver='warn',\n",
       "          tol=0.0001, verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "classifier.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "y_test_pred = classifier.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      1.00      1.00        16\n",
      "           1       1.00      0.76      0.87        17\n",
      "           2       0.75      1.00      0.86        12\n",
      "\n",
      "   micro avg       0.91      0.91      0.91        45\n",
      "   macro avg       0.92      0.92      0.91        45\n",
      "weighted avg       0.93      0.91      0.91        45\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(metrics.classification_report(y_test, y_test_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([16, 17, 12])"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.bincount(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[16,  0,  0],\n",
       "       [ 0, 13,  4],\n",
       "       [ 0,  0, 12]])"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metrics.confusion_matrix(y_test, y_test_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[16,  0,  0],\n",
       "       [ 0, 12,  5],\n",
       "       [ 0,  0, 12]])"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "classifier = tree.DecisionTreeClassifier()\n",
    "classifier.fit(X_train, y_train)\n",
    "y_test_pred = classifier.predict(X_test)\n",
    "metrics.confusion_matrix(y_test, y_test_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[16,  0,  0],\n",
       "       [ 0, 15,  2],\n",
       "       [ 0,  0, 12]])"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "classifier = neighbors.KNeighborsClassifier()\n",
    "classifier.fit(X_train, y_train)\n",
    "y_test_pred = classifier.predict(X_test)\n",
    "metrics.confusion_matrix(y_test, y_test_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[16,  0,  0],\n",
       "       [ 0, 15,  2],\n",
       "       [ 0,  0, 12]])"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "classifier = svm.SVC()\n",
    "classifier.fit(X_train, y_train)\n",
    "y_test_pred = classifier.predict(X_test)\n",
    "metrics.confusion_matrix(y_test, y_test_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[16,  0,  0],\n",
       "       [ 0, 16,  1],\n",
       "       [ 0,  0, 12]])"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "classifier = ensemble.RandomForestClassifier()\n",
    "classifier.fit(X_train, y_train)\n",
    "y_test_pred = classifier.predict(X_test)\n",
    "metrics.confusion_matrix(y_test, y_test_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "train_size_vec = np.linspace(0.1, 0.9, 30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "classifiers = [tree.DecisionTreeClassifier,\n",
    "               neighbors.KNeighborsClassifier,\n",
    "               svm.SVC,\n",
    "               ensemble.RandomForestClassifier\n",
    "              ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "cm_diags = np.zeros((3, len(train_size_vec), len(classifiers)), dtype=float)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2179: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.\n",
      "  FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n",
      "/Users/rob/miniconda3/envs/py3.6/lib/python3.6/site-packages/sklearn/ensemble/forest.py:246: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "for n, train_size in enumerate(train_size_vec):\n",
    "    X_train, X_test, y_train, y_test = \\\n",
    "        model_selection.train_test_split(iris.data, iris.target, train_size=train_size)\n",
    "\n",
    "    for m, Classifier in enumerate(classifiers): \n",
    "        classifier = Classifier()\n",
    "        classifier.fit(X_train, y_train)\n",
    "        y_test_pred = classifier.predict(X_test)\n",
    "        cm_diags[:, n, m] = metrics.confusion_matrix(y_test, y_test_pred).diagonal()\n",
    "        cm_diags[:, n, m] /= np.bincount(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x216 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(1, len(classifiers), figsize=(12, 3))\n",
    "\n",
    "for m, Classifier in enumerate(classifiers): \n",
    "    axes[m].plot(train_size_vec, cm_diags[2, :, m], label=iris.target_names[2])\n",
    "    axes[m].plot(train_size_vec, cm_diags[1, :, m], label=iris.target_names[1])\n",
    "    axes[m].plot(train_size_vec, cm_diags[0, :, m], label=iris.target_names[0])\n",
    "    axes[m].set_title(type(Classifier()).__name__)\n",
    "    axes[m].set_ylim(0, 1.1)\n",
    "    axes[m].set_xlim(0.1, 0.9)\n",
    "    axes[m].set_ylabel(\"classification accuracy\")\n",
    "    axes[m].set_xlabel(\"training size ratio\")\n",
    "    axes[m].legend(loc=4)\n",
    "\n",
    "fig.tight_layout()\n",
    "fig.savefig(\"ch15-classification-comparison.pdf\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Clustering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "X, y = iris.data, iris.target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "np.random.seed(123)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "n_clusters = 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "c = cluster.KMeans(n_clusters=n_clusters)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KMeans(algorithm='auto', copy_x=True, init='k-means++', max_iter=300,\n",
       "    n_clusters=3, n_init=10, n_jobs=None, precompute_distances='auto',\n",
       "    random_state=None, tol=0.0001, verbose=0)"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "c.fit(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "y_pred = c.predict(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 0, 0, 0, 0, 0, 0],\n",
       "      dtype=int32)"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred[::8]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2])"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y[::8]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "idx_0, idx_1, idx_2 = (np.where(y_pred == n) for n in range(3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "y_pred[idx_0], y_pred[idx_1], y_pred[idx_2] = 2, 0, 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2],\n",
       "      dtype=int32)"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred[::8]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[50,  0,  0],\n",
       "       [ 0, 48,  2],\n",
       "       [ 0, 14, 36]])"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metrics.confusion_matrix(y, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 16 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "N = X.shape[1]\n",
    "\n",
    "fig, axes = plt.subplots(N, N, figsize=(12, 12), sharex=True, sharey=True)\n",
    "\n",
    "colors = [\"coral\", \"blue\", \"green\"]\n",
    "markers = [\"^\", \"v\", \"o\"]\n",
    "for m in range(N):\n",
    "    for n in range(N):\n",
    "        for p in range(n_clusters):\n",
    "            mask = y_pred == p\n",
    "            axes[m, n].scatter(X[:, m][mask], X[:, n][mask],\n",
    "                               marker=markers[p], s=30, \n",
    "                               color=colors[p], alpha=0.25)\n",
    "\n",
    "        for idx in np.where(y != y_pred):\n",
    "            axes[m, n].scatter(X[idx, m], X[idx, n],\n",
    "                               marker=\"s\", s=30, \n",
    "                               edgecolor=\"red\", \n",
    "                               facecolor=(1,1,1,0))\n",
    "            \n",
    "            \n",
    "    axes[N-1, m].set_xlabel(iris.feature_names[m], fontsize=16)\n",
    "    axes[m, 0].set_ylabel(iris.feature_names[m], fontsize=16)\n",
    "fig.tight_layout()\n",
    "fig.savefig(\"ch15-clustering.pdf\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Versions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%reload_ext version_information"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table><tr><th>Software</th><th>Version</th></tr><tr><td>Python</td><td>3.6.8 64bit [GCC 4.2.1 Compatible Clang 4.0.1 (tags/RELEASE_401/final)]</td></tr><tr><td>IPython</td><td>7.5.0</td></tr><tr><td>OS</td><td>Darwin 18.2.0 x86_64 i386 64bit</td></tr><tr><td>sklearn</td><td>0.20.3</td></tr><tr><td>numpy</td><td>1.16.3</td></tr><tr><td>matplotlib</td><td>3.0.3</td></tr><tr><td>seaborn</td><td>0.9.0</td></tr><tr><td colspan='2'>Mon May 06 15:44:47 2019 JST</td></tr></table>"
      ],
      "text/latex": [
       "\\begin{tabular}{|l|l|}\\hline\n",
       "{\\bf Software} & {\\bf Version} \\\\ \\hline\\hline\n",
       "Python & 3.6.8 64bit [GCC 4.2.1 Compatible Clang 4.0.1 (tags/RELEASE\\_401/final)] \\\\ \\hline\n",
       "IPython & 7.5.0 \\\\ \\hline\n",
       "OS & Darwin 18.2.0 x86\\_64 i386 64bit \\\\ \\hline\n",
       "sklearn & 0.20.3 \\\\ \\hline\n",
       "numpy & 1.16.3 \\\\ \\hline\n",
       "matplotlib & 3.0.3 \\\\ \\hline\n",
       "seaborn & 0.9.0 \\\\ \\hline\n",
       "\\hline \\multicolumn{2}{|l|}{Mon May 06 15:44:47 2019 JST} \\\\ \\hline\n",
       "\\end{tabular}\n"
      ],
      "text/plain": [
       "Software versions\n",
       "Python 3.6.8 64bit [GCC 4.2.1 Compatible Clang 4.0.1 (tags/RELEASE_401/final)]\n",
       "IPython 7.5.0\n",
       "OS Darwin 18.2.0 x86_64 i386 64bit\n",
       "sklearn 0.20.3\n",
       "numpy 1.16.3\n",
       "matplotlib 3.0.3\n",
       "seaborn 0.9.0\n",
       "Mon May 06 15:44:47 2019 JST"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%version_information sklearn, numpy, matplotlib, seaborn"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "py3.6",
   "language": "python",
   "name": "py3.6"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}