{
 "metadata": {
  "name": ""
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "Predicting NFL Field Goal Percentages"
     ]
    },
    {
     "cell_type": "heading",
     "level": 6,
     "metadata": {},
     "source": [
      "*Disclaimer: All results and conclusions are subject to change*"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "This is a variation of the analysis done by [Greg Reda's](http://www.gregreda.com/2013/12/26/three-pointers-after-offensive-rebounds/) 3-point analysis for NCAA Basketball. Instead of NCAA basketball 3-pointers, I'm gonna look at field goal percentages in the NFL.  The overall purpose of this post is to learn more about Python's [pyMC3](https://github.com/pymc-devs/pymc/tree/pymc3) and [scikit-learn](http://scikit-learn.org/stable/).  The work based on scikit-learn is based on work done by [Trey Causey](https://twitter.com/treycausey) at [the spread](http://thespread.us/).  Here we will examine a [Random Forests Classifer](http://scikit-learn.org/dev/modules/ensemble.html) to predict field goal performance."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import numpy as np\n",
      "import scipy.stats as stats\n",
      "import pandas as pd\n",
      "import pandas.io.sql as psql\n",
      "import mysql.connector as sql\n",
      "\n",
      "%matplotlib inline\n",
      "import matplotlib.pyplot as plt\n",
      "\n",
      "# from mpld3 import enable_notebook\n",
      "# enable_notebook()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 12
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "I will be using the NFL data set from [Armchair Analysis](http://armchairanalysis.com/).  Their data is conveniently broken up into categories enabling quick analysis studies, they also provide scripts for setting up and importing the data into a SQL database.  The following code loads the data from a SQL database to a pandas DataFrame object using *pandas.io.sql*."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import pickle\n",
      "with open('sql.config','rb') as fp:\n",
      "    config = pickle.load(fp)\n",
      "\n",
      "con = sql.connect(**config)\n",
      "df = psql.frame_query('''\n",
      "    SELECT fgxp.DIST, fgxp.GOOD, games.TEMP, games.STAD FROM fgxp, core, games \n",
      "    WHERE core.PID=fgxp.PID AND games.GID = core.GID AND fgxp.FGXP='FG';''', con)\n",
      "\n",
      "con.close()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 13
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "The data for temperature was loaded into the SQL database as character string, convert it to a number."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "df.TEMP = df.TEMP.astype(float)\n",
      "df.head()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th></th>\n",
        "      <th>DIST</th>\n",
        "      <th>GOOD</th>\n",
        "      <th>TEMP</th>\n",
        "      <th>STAD</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>0</th>\n",
        "      <td> 43</td>\n",
        "      <td> Y</td>\n",
        "      <td> 79</td>\n",
        "      <td> Georgia Dome</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>1</th>\n",
        "      <td> 44</td>\n",
        "      <td> Y</td>\n",
        "      <td> 79</td>\n",
        "      <td> Georgia Dome</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>2</th>\n",
        "      <td> 24</td>\n",
        "      <td> Y</td>\n",
        "      <td> 79</td>\n",
        "      <td> Georgia Dome</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>3</th>\n",
        "      <td> 44</td>\n",
        "      <td> Y</td>\n",
        "      <td> 79</td>\n",
        "      <td> Georgia Dome</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>4</th>\n",
        "      <td> 48</td>\n",
        "      <td> Y</td>\n",
        "      <td> 79</td>\n",
        "      <td> Georgia Dome</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 14,
       "text": [
        "   DIST GOOD  TEMP          STAD\n",
        "0    43    Y    79  Georgia Dome\n",
        "1    44    Y    79  Georgia Dome\n",
        "2    24    Y    79  Georgia Dome\n",
        "3    44    Y    79  Georgia Dome\n",
        "4    48    Y    79  Georgia Dome"
       ]
      }
     ],
     "prompt_number": 14
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Next lets add the elevation data extracted using the code in another [notebook](http://nbviewer.ipython.org/github/jgbos/iPython-Notebooks/blob/master/Extracting%20Altitude%20of%20NFL%20Stadiums.ipynb).  "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "df_elevation = pd.read_csv('stadium_with_altitude.csv', index_col=0)\n",
      "df_elevation.head()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th></th>\n",
        "      <th>STAD</th>\n",
        "      <th>lat</th>\n",
        "      <th>lng</th>\n",
        "      <th>alt</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>0</th>\n",
        "      <td>             Georgia Dome</td>\n",
        "      <td> 33.757690</td>\n",
        "      <td>-84.400829</td>\n",
        "      <td> 303.163727</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>1</th>\n",
        "      <td> Cleveland Browns Stadium</td>\n",
        "      <td> 41.506483</td>\n",
        "      <td>-81.700040</td>\n",
        "      <td> 183.638306</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>2</th>\n",
        "      <td>            Texas Stadium</td>\n",
        "      <td> 32.841068</td>\n",
        "      <td>-96.910899</td>\n",
        "      <td> 133.471329</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>3</th>\n",
        "      <td>            Lambeau Field</td>\n",
        "      <td> 44.501357</td>\n",
        "      <td>-88.060755</td>\n",
        "      <td> 190.276779</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>4</th>\n",
        "      <td>        Arrowhead Stadium</td>\n",
        "      <td> 39.048939</td>\n",
        "      <td>-94.483916</td>\n",
        "      <td> 256.770477</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 16,
       "text": [
        "                       STAD        lat        lng         alt\n",
        "0              Georgia Dome  33.757690 -84.400829  303.163727\n",
        "1  Cleveland Browns Stadium  41.506483 -81.700040  183.638306\n",
        "2             Texas Stadium  32.841068 -96.910899  133.471329\n",
        "3             Lambeau Field  44.501357 -88.060755  190.276779\n",
        "4         Arrowhead Stadium  39.048939 -94.483916  256.770477"
       ]
      }
     ],
     "prompt_number": 16
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Now extract the elevation for every field goal attempt."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def getAlt(d):\n",
      "    c = df_elevation.STAD==d\n",
      "    if np.any(c): \n",
      "        return df_elevation[df_elevation.STAD==d]['alt'].iloc[0]\n",
      "    return np.nan\n",
      "\n",
      "df['ALT'] = df.STAD.map(getAlt)\n",
      "df.head()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th></th>\n",
        "      <th>DIST</th>\n",
        "      <th>GOOD</th>\n",
        "      <th>TEMP</th>\n",
        "      <th>STAD</th>\n",
        "      <th>ALT</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>0</th>\n",
        "      <td> 43</td>\n",
        "      <td> Y</td>\n",
        "      <td> 79</td>\n",
        "      <td> Georgia Dome</td>\n",
        "      <td> 303.163727</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>1</th>\n",
        "      <td> 44</td>\n",
        "      <td> Y</td>\n",
        "      <td> 79</td>\n",
        "      <td> Georgia Dome</td>\n",
        "      <td> 303.163727</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>2</th>\n",
        "      <td> 24</td>\n",
        "      <td> Y</td>\n",
        "      <td> 79</td>\n",
        "      <td> Georgia Dome</td>\n",
        "      <td> 303.163727</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>3</th>\n",
        "      <td> 44</td>\n",
        "      <td> Y</td>\n",
        "      <td> 79</td>\n",
        "      <td> Georgia Dome</td>\n",
        "      <td> 303.163727</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>4</th>\n",
        "      <td> 48</td>\n",
        "      <td> Y</td>\n",
        "      <td> 79</td>\n",
        "      <td> Georgia Dome</td>\n",
        "      <td> 303.163727</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 17,
       "text": [
        "   DIST GOOD  TEMP          STAD         ALT\n",
        "0    43    Y    79  Georgia Dome  303.163727\n",
        "1    44    Y    79  Georgia Dome  303.163727\n",
        "2    24    Y    79  Georgia Dome  303.163727\n",
        "3    44    Y    79  Georgia Dome  303.163727\n",
        "4    48    Y    79  Georgia Dome  303.163727"
       ]
      }
     ],
     "prompt_number": 17
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "MCMC Simulation For Distance and Temperature"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Now lets partition the data for doing an MCMC study of the field goal percentages for distances and temperature.  Lets look at distances less than 40 yards and define cold temperatures as those less than 40 degrees."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "convert = lambda x: True if x == 'Y' else False\n",
      "\n",
      "normal = (df.TEMP >= 40) & (df.DIST > 40) & df.TEMP.notnull()\n",
      "normal = df[normal].GOOD.apply(convert)\n",
      "\n",
      "cold = (df.TEMP < 40) & (df.DIST > 40) & df.TEMP.notnull()\n",
      "cold = df[cold].GOOD.apply(convert)\n",
      "\n",
      "print normal.mean(), normal.std()\n",
      "print cold.mean(), cold.std()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "0.659776992113 0.47384847452\n",
        "0.608137044968 0.488689847026\n"
       ]
      }
     ],
     "prompt_number": 18
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "It looks like there is a small difference in the mean, but not significant. Yet, these are just observations of a [Bernoulli Trial](http://en.wikipedia.org/wiki/Bernoulli_distribution).  To estimate the underlying probability of making a field goal in normal or cold conditions will execute an MCMC simulation.  The simulation will provide an estimate of the probability of making a field goal in normal versus cold conditions."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import pymc as pm"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 19
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "with pm.Model() as model:\n",
      "    p_normal = pm.Uniform(\"p_normal\",lower=0.2, upper=0.8)\n",
      "    p_cold = pm.Uniform(\"p_cold\",lower=0.4, upper=0.8)\n",
      "    \n",
      "    # scraped observations\n",
      "    obs_normal = pm.Bernoulli(\"obs_normal\", p_normal, observed=normal.astype(int))\n",
      "    obs_cold = pm.Bernoulli(\"obs_cold\", p_cold, observed=cold.astype(int))\n",
      "    \n",
      "    start = pm.find_MAP()\n",
      "    m = pm.NUTS(state=start)\n",
      "    trace = pm.sample(2000, m, start=start, progressbar=False)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "/Users/jgoodwin/anaconda/lib/python2.7/site-packages/theano/scan_module/scan_perform_ext.py:85: RuntimeWarning: numpy.ndarray size changed, may indicate binary incompatibility\n",
        "  from scan_perform.scan_perform import *\n"
       ]
      }
     ],
     "prompt_number": 20
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "p_normal_samples = trace['p_normal'][:]\n",
      "p_cold_samples=  trace['p_cold'][:]\n",
      "delta_samples = trace['p_normal']-trace['p_cold']"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 21
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plt.figure(figsize=(10, 10))\n",
      "\n",
      "ax = plt.subplot(311)\n",
      "plt.hist(p_normal_samples, histtype='stepfilled', bins=50, normed=True, color='#E41A1C', label='Warm')\n",
      "plt.legend()\n",
      "plt.vlines(normal.mean(), 0, 300, linestyles='--')\n",
      "plt.xticks(np.arange(0.5, 0.8, 0.05))\n",
      "plt.xlim(0.5, 0.8)\n",
      "plt.ylim(0, 80)\n",
      "plt.grid(False)\n",
      "\n",
      "ax = plt.subplot(312)\n",
      "plt.hist(p_cold_samples, histtype='stepfilled', bins=50, normed=True, color='#4DAF4A', label='Cold')\n",
      "plt.vlines(cold.mean(), 0, 300, linestyles='--')\n",
      "plt.legend()\n",
      "plt.xticks(np.arange(0.5, 0.8, 0.05))\n",
      "plt.xlim(0.5, 0.8)\n",
      "plt.ylim(0, 60)\n",
      "plt.grid(False)\n",
      "\n",
      "ax = plt.subplot(313)\n",
      "plt.hist(delta_samples, histtype='stepfilled', bins=50, normed=True, color='#377EB8', label='Delta')\n",
      "plt.vlines(0, 0, 300, linestyles='--', label='$H_0$ (No difference)')\n",
      "plt.legend()\n",
      "plt.xticks(np.arange(-0.1, 0.201, 0.1))\n",
      "plt.xlim(-0.1, 0.2)\n",
      "plt.ylim(0, 60)\n",
      "plt.grid(False);"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAlYAAAJPCAYAAAC+fJpMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
       "text": [
        "<matplotlib.figure.Figure at 0x110b4f4d0>"
       ]
      }
     ],
     "prompt_number": 29
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "print (\"Field Goals less than 40 yards are less successful in temperatures < 40 degrees \"\n",
      "       \"in {0:.1f}% of simulations\").format((delta_samples > 0).mean() * 100)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Field Goals less than 40 yards are less successful in temperatures < 40 degrees in 98.8% of simulations\n"
       ]
      }
     ],
     "prompt_number": 23
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "In approximately 99% of the simulations, field goals at colder temperatures were less likely to have been made."
     ]
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Random Forest Classification"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Now lets look at using a  [Random Forests Classifer](http://scikit-learn.org/dev/modules/ensemble.html) to predict the probability of successfully making a field goal given distance, temperature, and elevation of the stadium as training features.  "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from sklearn.ensemble import RandomForestClassifier\n",
      "from sklearn.cross_validation import train_test_split\n",
      "from sklearn.metrics import classification_report, roc_curve, auc, precision_recall_curve"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 24
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "First lets start with just distance and temperature as a features for determining whether a field goal is made.  The approach below will use 60% of the data for training and 40% to evaluate performance.  Lets develop a model."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "convert = lambda x: 1 if x == 'Y' else 0\n",
      "data = df;\n",
      "data = data[data.TEMP.notnull()]\n",
      "data.TEMP = data.TEMP.astype(float)\n",
      "data.GOOD = data.GOOD.apply(convert)\n",
      "\n",
      "X = data[['DIST','TEMP']]\n",
      "y = data['GOOD']\n",
      "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=.4)\n",
      "\n",
      "# Set up the random forest model with parameters found via grid search\n",
      "clf = RandomForestClassifier(oob_score = True, criterion='gini', n_estimators=100,  min_samples_leaf=20, max_features=2)\n",
      "clf.fit(X_train, y_train)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 25,
       "text": [
        "RandomForestClassifier(bootstrap=True, compute_importances=None,\n",
        "            criterion='gini', max_depth=None, max_features=2,\n",
        "            min_density=None, min_samples_leaf=20, min_samples_split=2,\n",
        "            n_estimators=100, n_jobs=1, oob_score=True, random_state=None,\n",
        "            verbose=0)"
       ]
      }
     ],
     "prompt_number": 25
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def summary():\n",
      "    print 'Training set score: %.1f%%' % (clf.score(X_train, y_train)*100)\n",
      "    print 'Test set score: %.1f%%' %(clf.score(X_test, y_test)*100)\n",
      "    preds = clf.predict(X_test)\n",
      "    print classification_report(np.array(y_test), preds)\n",
      "    print 'Confusion matrix: '\n",
      "    print pd.crosstab(y_test, preds, rownames=[\"Actual\"], colnames=[\"Predicted\"])\n",
      "\n",
      "# summary()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 26
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def roc():\n",
      "    full_probs = clf.predict_proba(X_test)\n",
      "    fpr_full, tpr_full, thresholds_full = roc_curve(y_test, full_probs[:, 1])\n",
      "    roc_auc_full = auc(fpr_full, tpr_full)\n",
      "    \n",
      "    plt.figure(figsize=(6, 5))\n",
      "    plt.plot(fpr_full, tpr_full, label='Full random forest ROC curve (area = %0.2f)' % roc_auc_full)\n",
      "    plt.plot([0, 1], [0, 1], 'k--')\n",
      "    plt.xlim([0.0, 1.0])\n",
      "    plt.ylim([0.0, 1.0])\n",
      "    plt.xlabel('False Positive Rate')\n",
      "    plt.ylabel('True Positive Rate')\n",
      "    plt.title('Receiver operating characteristic curve')\n",
      "    plt.legend(loc=\"lower right\");\n",
      "roc()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAYYAAAFRCAYAAABwuwpyAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x11009dd90>"
       ]
      }
     ],
     "prompt_number": 27
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Above is a [ROC curve](http://en.wikipedia.org/wiki/Receiver_operating_characteristic) illustrating our model's performance against the test data. It doesn't show the best performance, but clearly doing better than flipping a coin."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "h = 5  # step size in the mesh\n",
      "x_min, x_max = X.DIST.min() - .5, X.DIST.max() + .5\n",
      "y_min, y_max = X.TEMP.min() - .5, X.TEMP.max() + .5\n",
      "xx, yy = np.meshgrid(np.arange(x_min, x_max, h), np.arange(y_min, y_max, h))\n",
      "s = xx.ravel().shape\n",
      "cm = plt.cm.RdBu\n",
      "\n",
      "v = None\n",
      "for estimator in clf.estimators_:\n",
      "    e = estimator.predict_proba(np.c_[xx.ravel(), yy.ravel()])[:, 1]    \n",
      "    if v is None:\n",
      "        v = e.reshape((len(e),1))\n",
      "    else:\n",
      "        v = np.hstack((v, e.reshape((len(e),1))))\n",
      "Z = v.mean(axis=1)\n",
      "p1 = np.percentile(v, 2.5, axis=1)\n",
      "p2 = np.percentile(v, 97.5, axis=1)\n",
      "P = p2-p1\n",
      "\n",
      "\n",
      "fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(12,4))\n",
      "\n",
      "plt.subplot(121)\n",
      "Z = Z.reshape(xx.shape)\n",
      "plt.contourf(xx, yy, Z, cmap=cm, alpha=.8, levels=np.arange(0,1.1,0.1))\n",
      "plt.title('Prediction')\n",
      "plt.ylabel('Temperature (${\\ }^{\\circ} F$)')\n",
      "plt.xlabel('Yards From Scrimmage')\n",
      "\n",
      "plt.subplot(122)\n",
      "P = P.reshape(xx.shape)\n",
      "co = plt.contourf(xx, yy, P, cmap=cm, alpha=.8, levels=np.arange(0,1.1,0.1))\n",
      "plt.xlabel('Yards From Scrimmage')\n",
      "plt.title('Width of Confidence Interval')\n",
      "\n",
      "fig.subplots_adjust(right=0.8)\n",
      "cbar_ax = fig.add_axes([0.81, 0.15, 0.02, 0.7])\n",
      "fig.colorbar(co, cax=cbar_ax);\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAqoAAAEZCAYAAABfMwcsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/gb42yA02CtSDgAAAABJRU5ErkJggg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x110b3f150>"
       ]
      }
     ],
     "prompt_number": 28
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "The above figure illustrates the predicted probability of making (left) a Field Goal at a given temperature versus a given distance and the width of the 95% confidence interval estimated from the distribution of estimator outputs (left).  A couple quick observations: as expected, smaller distance mean higher likelihood the field goal will be made; the confidence interval width is smaller in regions where more data is available.  So does temperature impact the likelihood of making a field goal?  It appears to be a minor impact, right around 40 degrees there is slight change in the probability.  From our MCMC simulation this difference was only a couple percentage points, and the figure above appears to illustrate the same thing.  Looking at the 90th-percentile probability, there is a shift of about 5 yards for the same probability.  So temperature has an impact, but very minor.  Below is the output of the feature importance estimated from the classifier, obviously distance is more important than temperature."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "print \"Distance: %f, Temperature: %f\" %(clf.feature_importances_[0],clf.feature_importances_[1])"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Distance: 0.730580, Temperature: 0.269420\n"
       ]
      }
     ],
     "prompt_number": 30
    },
    {
     "cell_type": "heading",
     "level": 3,
     "metadata": {},
     "source": [
      "Include Altitude in Our Model"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "convert = lambda x: 1 if x == 'Y' else 0\n",
      "data = df;\n",
      "data = data[data.TEMP.notnull()]\n",
      "data.TEMP = data.TEMP.astype(float)\n",
      "data.GOOD = data.GOOD.apply(convert)\n",
      "\n",
      "X = data[['DIST','TEMP','ALT']]\n",
      "y = data['GOOD']\n",
      "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=.4)\n",
      "\n",
      "# Set up the random forest model with parameters found via grid search\n",
      "clf = RandomForestClassifier(oob_score = True, criterion='entropy', n_estimators=100,  min_samples_leaf=15)\n",
      "clf.fit(X_train, y_train)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 31,
       "text": [
        "RandomForestClassifier(bootstrap=True, compute_importances=None,\n",
        "            criterion='entropy', max_depth=None, max_features='auto',\n",
        "            min_density=None, min_samples_leaf=15, min_samples_split=2,\n",
        "            n_estimators=100, n_jobs=1, oob_score=True, random_state=None,\n",
        "            verbose=0)"
       ]
      }
     ],
     "prompt_number": 31
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "roc()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAYYAAAFRCAYAAABwuwpyAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x104829450>"
       ]
      }
     ],
     "prompt_number": 32
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "The ROC curve above does not show much improvement over the 2 feature classifier.  Lets look at the predicted probabilities and confidence interval with altitude included in our model."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "h = 5  # step size in the mesh\n",
      "x_min, x_max = X.DIST.min() - .5, X.DIST.max() + .5\n",
      "y_min, y_max = X.TEMP.min() - .5, X.TEMP.max() + .5\n",
      "z_min, z_max = X.ALT.min() - .5, X.ALT.max() + .5\n",
      "xx, yy = np.meshgrid(np.arange(x_min, x_max, h), np.arange(y_min, y_max, h))\n",
      "s = xx.ravel().shape\n",
      "cm = plt.cm.RdBu\n",
      "\n",
      "def predict(z):\n",
      "    v = None\n",
      "    for estimator in clf.estimators_:\n",
      "        e = estimator.predict_proba(np.c_[xx.ravel(), yy.ravel(), np.repeat(z,s)])[:, 1]    \n",
      "        if v is None:\n",
      "            v = e.reshape((len(e),1))\n",
      "        else:\n",
      "            v = np.hstack((v, e.reshape((len(e),1))))\n",
      "    m = v.mean(axis=1)\n",
      "    p1 = np.percentile(v, 2.5, axis=1)\n",
      "    p2 = np.percentile(v, 97.5, axis=1)\n",
      "    p = p2-p1\n",
      "    return m, p\n",
      "\n",
      "\n",
      "fig, axes = plt.subplots(nrows=3, ncols=2, figsize=(10,8))\n",
      "\n",
      "plt.subplot(321)\n",
      "Z,P = predict(5.0)\n",
      "# Z = clf.predict_proba(np.c_[xx.ravel(), yy.ravel(), np.repeat(5.0,s)])[:, 1]\n",
      "Z = Z.reshape(xx.shape)\n",
      "plt.contourf(xx, yy, Z, cmap=cm, alpha=.8, levels=np.arange(0,1.1,0.1))\n",
      "plt.title('5 Meters Alititude Prediction')\n",
      "plt.ylabel('Temperature (${\\ }^{\\circ} F$)')\n",
      "\n",
      "plt.subplot(322)\n",
      "P = P.reshape(xx.shape)\n",
      "plt.contourf(xx, yy, P, cmap=cm, alpha=.8, levels=np.arange(0,1.1,0.1))\n",
      "plt.title('5 Meters Alitude Confidence Interval Width')\n",
      "\n",
      "\n",
      "\n",
      "plt.subplot(323)\n",
      "Z,P = predict(500.0)\n",
      "# Z = clf.predict_proba(np.c_[xx.ravel(), yy.ravel(), np.repeat(500.0,s)])[:, 1]\n",
      "Z = Z.reshape(xx.shape)\n",
      "plt.contourf(xx, yy, Z, cmap=cm, alpha=.8, levels=np.arange(0,1.1,0.1))\n",
      "plt.title('500 Meters Alititude Prediction')\n",
      "plt.ylabel('Temperature (${\\ }^{\\circ} F$)')\n",
      "\n",
      "plt.subplot(324)\n",
      "P = P.reshape(xx.shape)\n",
      "plt.contourf(xx, yy, P, cmap=cm, alpha=.8, levels=np.arange(0,1.1,0.1))\n",
      "plt.title('500 Meters Alitude Confidence Interval Width')\n",
      "\n",
      "\n",
      "plt.subplot(325)\n",
      "Z,P = predict(1500.0)\n",
      "# Z = clf.predict_proba(np.c_[xx.ravel(), yy.ravel(), np.repeat(1500.0,s)])[:, 1]\n",
      "Z = Z.reshape(xx.shape)\n",
      "plt.contourf(xx, yy, Z, cmap=cm, alpha=.8, levels=np.arange(0,1.1,0.1))\n",
      "plt.title('1500 Meters Alititude Prediction')\n",
      "plt.xlabel('Yards From Scrimmage')\n",
      "plt.ylabel('Temperature (${\\ }^{\\circ} F$)')\n",
      "\n",
      "plt.subplot(326)\n",
      "P = P.reshape(xx.shape)\n",
      "plt.contourf(xx, yy, P, cmap=cm, alpha=.8, levels=np.arange(0,1.1,0.1))\n",
      "plt.title('1500 Meters Alitude Confidence Interval Width')\n",
      "plt.xlabel('Yards From Scrimmage')\n",
      "\n",
      "fig.subplots_adjust(right=.8)\n",
      "cbar_ax = fig.add_axes([1, 0.15, 0.02, 0.7])\n",
      "fig.colorbar(co, cax=cbar_ax)\n",
      "\n",
      "plt.tight_layout()\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAvUAAAI7CAYAAABsoGRJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nh9SonmIAAAAASUVORK5CYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0x1102c3a50>"
       ]
      }
     ],
     "prompt_number": 23
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Not too much to take from these results.  Higher altitudes have some impact on the probability. A pattern shows up at low altitude for temperature, otherwise at higher altitude temperature doesn't really play a role (especially when accounting for the confidence interval).  Probably the most notable change in probability across altitude is in the 80th-percentile probabilities at lower temperatures.  It looks like there is an increase in probability at lower temperature from lower altitudes.  "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plt.figure(figsize=(12,6))\n",
      "plt.bar([0,1,2],clf.feature_importances_, align='center')\n",
      "plt.xticks([0,1,2],['Distance','Temperature','Altitude']);"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAr4AAAFwCAYAAABejXgsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nCcIXAIAk/B+F2i1/x47wRgAAAABJRU5ErkJggg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x10fc1e0d0>"
       ]
      }
     ],
     "prompt_number": 21
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "References"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "1. [Going for Three: Predicting the Likelihood of Field Goal Success with Logistic Regression](http://www.sloansportsconference.com/wp-content/uploads/2013/Slides/RP/Going%20for%20Three_Predicting%20the%20likelihood%20of%20field%20goal%20success%20with%20logistic%20regression.pdf)\n",
      "2. [Temperature and Field Goals](http://www.advancednflstats.com/2012/01/temperature-and-field-goals.html)\n",
      "3. [Altitude and Field Goals](http://www.advancednflstats.com/2013/01/altitude-and-field-goals.html)\n",
      "4. [3-pointers after offensive rebounds](http://www.gregreda.com/2013/12/26/three-pointers-after-offensive-rebounds/)\n",
      "5. [the spread](http://thespread.us/)"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from IPython.core.display import HTML\n",
      "def css_styling():\n",
      "    styles = open(\"style.css\", \"r\").read()\n",
      "    return HTML(styles)\n",
      "css_styling()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<style>\n",
        "\n",
        "   .cell {\n",
        "       width: inherit !important;\n",
        "       background-color: #f3f3f3 !important;\n",
        "       margin-left: 2% !important;\n",
        "   }\n",
        "\n",
        "   /* block fixes margin on input boxes */\n",
        "   /* in firefox */\n",
        "   .input.hbox {\n",
        "        width: 80% !important;\n",
        "        display: block !important;\n",
        "    }\n",
        "    \n",
        "   .input_area { \n",
        "        background-color: white !important; \n",
        "    }\n",
        "\n",
        "    .output_area pre {\n",
        "        font-family: \"Source Code Pro\", source-code-pro, Consolas, monospace;\n",
        "        border: 0px;        \n",
        "    }\n",
        "\n",
        "    div.output_text {\n",
        "        font-family: \"Source Code Pro\", source-code-pro, Consolas, monospace; \n",
        "    }\n",
        "\n",
        "    div.text_cell {\n",
        "        width: 80%;\n",
        "        text-align: left;\n",
        "    }\n",
        "    \n",
        "    div.prompt {\n",
        "        width: 0px;\n",
        "        visibility: hidden !important;\n",
        "    }\n",
        "    \n",
        "    .code_cell {\n",
        "        background-color: #f3f3f3;\n",
        "    }\n",
        "    \n",
        "    .highlight {\n",
        "        background-color: #ffffff;\n",
        "    }\n",
        "    \n",
        "    div.input_prompt {\n",
        "        visibility: hidden;\n",
        "        width: 0 !important;\n",
        "    }\n",
        "\n",
        "    div.text_cell_render {\n",
        "        font-family: \"Minion Pro\", \"minion-pro\", \"Charis SIL\", Palatino, serif !important;\n",
        "        font-size: 14pt !important;\n",
        "        line-height: 145% !important;\n",
        "        width: 80% !important;\n",
        "        text-align: left !important;\n",
        "        background-color: #f3f3f3 !important;\n",
        "    }\n",
        "\n",
        "    div.text_cell_render h1 {\n",
        "        display: block;\n",
        "        font-size: 28pt;\n",
        "        color: #3B3B3B;\n",
        "        margin-bottom: 0em;\n",
        "        margin-top: 0.5em;\n",
        "        \n",
        "    }\n",
        "    \n",
        "    .rendered_html li {\n",
        "        margin-bottom: .25em;\n",
        "        color: #3B3B3B;;\n",
        "    }\n",
        "    \n",
        "    div.text_cell_render h2:before {\n",
        "        content: \"\\2FFA\";\n",
        "        margin-right: 0.5em;\n",
        "        font-size: .5em;\n",
        "        vertical-align: baseline;\n",
        "        border-top: 1px;\n",
        "\n",
        "    }\n",
        "\n",
        "    .hiterm {\n",
        "        font-weight: 500;\n",
        "        color: #DC143C;\n",
        "    }\n",
        "\n",
        "   .text_cell_render h2 {\n",
        "        font-size: 20pt;\n",
        "        margin-bottom: 0em;\n",
        "        margin-top: 0.5em;\n",
        "        display: block;\n",
        "        color: #3B3B3B;\n",
        "    }\n",
        "\n",
        "    .MathJax_Display {\n",
        "        /*text-align: center !important;*/\n",
        "        margin-left: 2em !important;\n",
        "        margin-top: .5em !important;\n",
        "        margin-bottom: .5em !important;\n",
        "    }\n",
        "\n",
        "    .text_cell_render h3 {\n",
        "        font-size: 14pt;\n",
        "        font-weight: 600;\n",
        "        font-style: italic;\n",
        "        margin-bottom: -0.5em;\n",
        "        margin-top: -0.25em;\n",
        "        color: #3B3B3B;\n",
        "        text-indent: 2em;\n",
        "    }\n",
        "\n",
        "    .text_cell_render h5 {\n",
        "        font-weight: 300;\n",
        "        font-size: 14pt;\n",
        "        color: #4057A1;\n",
        "        font-style: italic;\n",
        "        margin-bottom: .5em;\n",
        "        margin-top: 0.5em;\n",
        "        display: block;\n",
        "    }\n",
        "    \n",
        "    .CodeMirror {\n",
        "        font-family: \"Source Code Pro\", source-code-pro, Consolas, monospace !important;\n",
        "        font-size: 10pt;\n",
        "        background:  #fffffe; /* #f0f8fb #e2eef9*/\n",
        "        border: 0px;\n",
        "    }\n",
        "    \n",
        "    .rendered_html {\n",
        "        margin-left: 2%;\n",
        "        \n",
        "    }\n",
        "\n",
        "    .rendered_html code {\n",
        "        font-family: \"Source Code Pro\", source-code-pro,Consolas,    monospace;\n",
        "        font-size: 85%;\n",
        "    }\n",
        "    \n",
        "   pre, code, kbd, samp { font-family: \"Source Code Pro\", source-code-pro, Consola, monospace; }\n",
        "    \n",
        "    .rendered_html p {\n",
        "        text-align: left;\n",
        "        color: #3B3B3B;\n",
        "        margin-bottom: .5em;\n",
        "\n",
        "    }\n",
        "\n",
        "       .rendered_html p+p { \n",
        "          text-indent: 1em;\n",
        "           margin-top: 0;\n",
        "       }\n",
        "    .rendered_html ol {\n",
        "        list-style: decimal;\n",
        "        /*margin: 1em 2em;*/\n",
        "    }\n",
        "\n",
        "    .rendered_html ol ol {\n",
        "        list-style: decimal;\n",
        "    }\n",
        "\n",
        "    .rendered_html ol ol ol {\n",
        "        list-style: decimal;\n",
        "    }\n",
        "\n",
        "    body{background-color:#f3f3f3;}\n",
        "\n",
        "    .rendered_html p.hangpar {\n",
        "        text-indent: 0;\n",
        "    }\n",
        "\n",
        "</style>"
       ],
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 11,
       "text": [
        "<IPython.core.display.HTML at 0x104824390>"
       ]
      }
     ],
     "prompt_number": 11
    }
   ],
   "metadata": {}
  }
 ]
}