P4DS4D; 19; Increasing Complexity.ipynb

{
 "metadata": {
  "name": "",
  "signature": "sha256:a0226d66f72ed641347016b8172fc2899d7f55c73b8d66e66067f49a1da9915e"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "Using Non-linear Transformations"
     ]
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Doing variable transformations"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from sklearn.datasets import load_boston\n",
      "from random import shuffle\n",
      "boston = load_boston()\n",
      "seed(0) # Creates a replicable shuffling\n",
      "new_index = range(boston.data.shape[0])\n",
      "shuffle(new_index) # shuffling the index\n",
      "X, y = boston.data[new_index], boston.target[new_index]\n",
      "print X.shape, y.shape, boston.feature_names"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "(506L, 13L) (506L,) ['CRIM' 'ZN' 'INDUS' 'CHAS' 'NOX' 'RM' 'AGE' 'DIS' 'RAD' 'TAX' 'PTRATIO'\n",
        " 'B' 'LSTAT']\n"
       ]
      }
     ],
     "prompt_number": 13
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "print boston.DESCR"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Boston House Prices dataset\n",
        "\n",
        "Notes\n",
        "------\n",
        "Data Set Characteristics:  \n",
        "\n",
        "    :Number of Instances: 506 \n",
        "\n",
        "    :Number of Attributes: 13 numeric/categorical predictive\n",
        "    \n",
        "    :Median Value (attribute 14) is usually the target\n",
        "\n",
        "    :Attribute Information (in order):\n",
        "        - CRIM     per capita crime rate by town\n",
        "        - ZN       proportion of residential land zoned for lots over 25,000 sq.ft.\n",
        "        - INDUS    proportion of non-retail business acres per town\n",
        "        - CHAS     Charles River dummy variable (= 1 if tract bounds river; 0 otherwise)\n",
        "        - NOX      nitric oxides concentration (parts per 10 million)\n",
        "        - RM       average number of rooms per dwelling\n",
        "        - AGE      proportion of owner-occupied units built prior to 1940\n",
        "        - DIS      weighted distances to five Boston employment centres\n",
        "        - RAD      index of accessibility to radial highways\n",
        "        - TAX      full-value property-tax rate per $10,000\n",
        "        - PTRATIO  pupil-teacher ratio by town\n",
        "        - B        1000(Bk - 0.63)^2 where Bk is the proportion of blacks by town\n",
        "        - LSTAT    % lower status of the population\n",
        "        - MEDV     Median value of owner-occupied homes in $1000's\n",
        "\n",
        "    :Missing Attribute Values: None\n",
        "\n",
        "    :Creator: Harrison, D. and Rubinfeld, D.L.\n",
        "\n",
        "This is a copy of UCI ML housing dataset.\n",
        "http://archive.ics.uci.edu/ml/datasets/Housing\n",
        "\n",
        "\n",
        "This dataset was taken from the StatLib library which is maintained at Carnegie Mellon University.\n",
        "\n",
        "The Boston house-price data of Harrison, D. and Rubinfeld, D.L. 'Hedonic\n",
        "prices and the demand for clean air', J. Environ. Economics & Management,\n",
        "vol.5, 81-102, 1978.   Used in Belsley, Kuh & Welsch, 'Regression diagnostics\n",
        "...', Wiley, 1980.   N.B. Various transformations are used in the table on\n",
        "pages 244-261 of the latter.\n",
        "\n",
        "The Boston house-price data has been used in many machine learning papers that address regression\n",
        "problems.   \n",
        "     \n",
        "**References**\n",
        "\n",
        "   - Belsley, Kuh & Welsch, 'Regression diagnostics: Identifying Influential Data and Sources of Collinearity', Wiley, 1980. 244-261.\n",
        "   - Quinlan,R. (1993). Combining Instance-Based and Model-Based Learning. In Proceedings on the Tenth International Conference of Machine Learning, 236-243, University of Massachusetts, Amherst. Morgan Kaufmann.\n",
        "   - many more! (see http://archive.ics.uci.edu/ml/datasets/Housing)\n",
        "\n"
       ]
      }
     ],
     "prompt_number": 14
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import pandas as pd\n",
      "df = pd.DataFrame(X,columns=boston.feature_names)\n",
      "df['target'] = y"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 15
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "scatter = df.plot(kind='scatter', x='LSTAT', y='target', c='r')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAYEAAAEPCAYAAACk43iMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x14bbf3c8>"
       ]
      }
     ],
     "prompt_number": 16
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import numpy as np\n",
      "from sklearn.feature_selection.univariate_selection import f_regression\n",
      "F, pval = f_regression(df['LSTAT'],y)\n",
      "print 'F score for the original feature %.1f' % F\n",
      "F, pval = f_regression(np.log(df['LSTAT']),y)\n",
      "print 'F score for the transformed feature %.1f' % F"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "F score for the original feature 601.6\n",
        "F score for the transformed feature 1000.2\n"
       ]
      }
     ],
     "prompt_number": 17
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Creating interactions between variables"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from sklearn.linear_model import LinearRegression\n",
      "from sklearn.cross_validation import cross_val_score\n",
      "from sklearn.cross_validation import KFold\n",
      "regression = LinearRegression(normalize=True)\n",
      "crossvalidation = KFold(n=X.shape[0], n_folds=10, shuffle=True, random_state=1)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 18
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "df = pd.DataFrame(X,columns=boston.feature_names)\n",
      "baseline = np.mean(cross_val_score(regression, df, y, scoring='r2', cv=crossvalidation, n_jobs=1))\n",
      "interactions = list()\n",
      "for feature_A in boston.feature_names:\n",
      "    for feature_B in boston.feature_names:\n",
      "        if feature_A > feature_B:\n",
      "            df['interaction'] = df[feature_A] * df[feature_B]\n",
      "            score = np.mean(cross_val_score(regression, df, y, scoring='r2', cv=crossvalidation, n_jobs=1))\n",
      "            if score > baseline:\n",
      "                interactions.append((feature_A, feature_B, round(score,3)))\n",
      "print 'Baseline R2: %.3f' % baseline\n",
      "print 'Top 10 interactions: %s' % sorted(interactions, key=lambda(x):x[2], reverse=True)[:10]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Baseline R2: 0.720\n",
        "Top 10 interactions: [('RM', 'LSTAT', 0.79), ('TAX', 'RM', 0.783), ('RM', 'RAD', 0.777), ('RM', 'PTRATIO', 0.768), ('RM', 'INDUS', 0.758), ('RM', 'NOX', 0.75), ('RM', 'AGE', 0.747), ('RM', 'B', 0.746), ('RM', 'DIS', 0.739), ('INDUS', 'DIS', 0.733)]\n"
       ]
      }
     ],
     "prompt_number": 19
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "colors = ['k' if v > np.mean(y) else 'w' for v in y]\n",
      "scatter = df.plot(kind='scatter', x='RM', y='LSTAT', c=colors)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEPCAYAAABP1MOPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x14b50cc0>"
       ]
      }
     ],
     "prompt_number": 20
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Adding polynomial features (x**2,y**2,xy)\n",
      "polyX = pd.DataFrame(X,columns=boston.feature_names)\n",
      "baseline = np.mean(cross_val_score(regression, polyX, y, scoring='mean_squared_error', cv=crossvalidation, n_jobs=1))\n",
      "improvements = [baseline]\n",
      "for feature_A in boston.feature_names: \n",
      "    polyX[feature_A+'^2'] = polyX[feature_A]**2\n",
      "    improvements.append(np.mean(cross_val_score(regression, polyX, y, scoring='mean_squared_error', cv=crossvalidation, n_jobs=1)))\n",
      "    for feature_B in boston.feature_names:\n",
      "        if feature_A > feature_B:\n",
      "            polyX[feature_A+'*'+feature_B] = polyX[feature_A] * polyX[feature_B]\n",
      "            improvements.append(np.mean(cross_val_score(regression, polyX, y, scoring='mean_squared_error', cv=crossvalidation, n_jobs=1)))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 21
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import matplotlib.pyplot as plt\n",
      "plt.figure() \n",
      "plt.plot(range(0,92),np.abs(improvements),'-')\n",
      "plt.xlabel('Polynomial features')\n",
      "plt.ylabel('Mean squared error')\n",
      "plt.show()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAYQAAAEPCAYAAABCyrPIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x14bbc438>"
       ]
      }
     ],
     "prompt_number": 22
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "print shape(polyX)\n",
      "crossvalidation = KFold(n=X.shape[0], n_folds=10, shuffle=True, random_state=1)\n",
      "print 'Mean squared error %.3f' % abs(np.mean(cross_val_score(regression, polyX, y, scoring='mean_squared_error', cv=crossvalidation, n_jobs=1)))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "(506, 104)\n",
        "Mean squared error 13.466"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n"
       ]
      }
     ],
     "prompt_number": 25
    },
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "Regularizing linear models"
     ]
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Relying on ridge regression"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from sklearn.grid_search import GridSearchCV\n",
      "from sklearn.linear_model import Ridge\n",
      "ridge = Ridge(normalize=True)\n",
      "search = GridSearchCV(estimator=ridge, param_grid={'alpha':np.logspace(-5,2,8)}, \n",
      "                                       scoring='mean_squared_error', n_jobs=1, refit=True, cv=10)\n",
      "search.fit(polyX,y)\n",
      "print 'Best parameters: %s' % search.best_params_\n",
      "print 'CV MSE of best parameters: %.3f' % abs(search.best_score_)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Best parameters: {'alpha': 0.001}\n",
        "CV MSE of best parameters: 12.385\n"
       ]
      }
     ],
     "prompt_number": 28
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Using the Lasso"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from sklearn.linear_model import Lasso\n",
      "lasso = Lasso(normalize=True)\n",
      "search = GridSearchCV(estimator=lasso, param_grid={'alpha':np.logspace(-5,2,8)}, \n",
      "                                       scoring='mean_squared_error', n_jobs=1, refit=True, cv=10)\n",
      "search.fit(polyX,y)\n",
      "print 'Best parameters: %s' % search.best_params_\n",
      "print 'CV MSE of best parameters: %.3f' % abs(search.best_score_)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Best parameters: {'alpha': 0.0001}\n",
        "CV MSE of best parameters: 12.644\n"
       ]
      }
     ],
     "prompt_number": 29
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Combining L1 & L2: Elasticnet"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from sklearn.linear_model import ElasticNet\n",
      "elastic = ElasticNet(normalize=True)\n",
      "search = GridSearchCV(estimator=elastic, param_grid={'alpha':np.logspace(-5,2,8), 'l1_ratio': [0.25, 0.5, 0.75]}, \n",
      "                                       scoring='mean_squared_error', n_jobs=1, refit=True, cv=10)\n",
      "# For l1_ratio = 0 the penalty is an L2 penalty. For l1_ratio = 1 it is an L1 penalty\n",
      "search.fit(polyX,y)\n",
      "print 'Best parameters: %s' % search.best_params_\n",
      "print 'CV MSE of best parameters: %.3f' % abs(search.best_score_)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Best parameters: {'alpha': 1.0000000000000001e-05, 'l1_ratio': 0.5}\n",
        "CV MSE of best parameters: 12.162\n"
       ]
      }
     ],
     "prompt_number": 31
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Leveraging regularization for feature selection"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "lasso = Lasso(normalize=True, alpha=0.01)\n",
      "lasso.fit(polyX,y)\n",
      "print polyX.columns[np.abs(lasso.coef_)>0.0001].values"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "['CRIM*CHAS' 'ZN*CRIM' 'ZN*CHAS' 'INDUS*DIS' 'CHAS*B' 'NOX^2' 'NOX*DIS'\n",
        " 'RM^2' 'RM*CRIM' 'RM*NOX' 'RM*PTRATIO' 'RM*B' 'RM*LSTAT' 'RAD*B' 'TAX*DIS'\n",
        " 'PTRATIO*NOX' 'LSTAT^2']\n"
       ]
      }
     ],
     "prompt_number": 32
    },
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "Fighting with Big Data: Stochastic Gradient Descent"
     ]
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Determining when there is too much data"
     ]
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Implementing classification and regression with SGD"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from sklearn.linear_model import SGDRegressor\n",
      "from sklearn.preprocessing import StandardScaler\n",
      "\n",
      "SGD = SGDRegressor(loss='squared_loss', penalty='l2', alpha=0.0001, l1_ratio=0.15, n_iter=2000)\n",
      "#  l1_ratio=0 corresponds to L2 penalty, l1_ratio=1 to L1\n",
      "# alpha constant that multiplies the regularization term. Defaults to 0.0001\n",
      "scaling = StandardScaler()\n",
      "scaling.fit(polyX)\n",
      "scaled_X = scaling.transform(polyX)\n",
      "print 'CV MSE: %.3f' % abs(np.mean(cross_val_score(SGD, scaled_X, y, scoring='mean_squared_error', cv=crossvalidation, n_jobs=1)))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "CV MSE: 12.802\n"
       ]
      }
     ],
     "prompt_number": 34
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from sklearn.metrics import mean_squared_error\n",
      "from sklearn.cross_validation import train_test_split\n",
      "\n",
      "X_train, X_test, y_train, y_test = train_test_split(scaled_X, y, test_size=0.20, random_state=2)\n",
      "\n",
      "SGD = SGDRegressor(loss='squared_loss', penalty='l2', alpha=0.0001, l1_ratio=0.15)\n",
      "improvements = list()\n",
      "for z in range(1000):\n",
      "    SGD.partial_fit(X_train, y_train)\n",
      "    improvements.append(mean_squared_error(y_test, SGD.predict(X_test)))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 35
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import matplotlib.pyplot as plt\n",
      "plt.subplot(1,2,1) \n",
      "plt.plot(range(1,11),np.abs(improvements[:10]),'o--')\n",
      "plt.xlabel('Partial fit initial iterations')\n",
      "plt.ylabel('Test set mean squared error')\n",
      "plt.subplot(1,2,2) \n",
      "plt.plot(range(100,1000,100),np.abs(improvements[100:1000:100]),'o--')\n",
      "plt.xlabel('Partial fit ending iterations')\n",
      "plt.show()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAYQAAAEPCAYAAABCyrPIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x14750e48>"
       ]
      }
     ],
     "prompt_number": 36
    }
   ],
   "metadata": {}
  }
 ]
}