{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#Decision Tree Classifiers\n",
    "<p>This Notebook gives an overview of DecisionTrees with scikit learn. In it are examples of:\n",
    "\n",
    "<ul>\n",
    "<li>Build a basic decision tree</li>\n",
    "<li>Plotting the decision surface of a decision tree</li>\n",
    "<li>Visualing the decision tree</li>\n",
    "<li>Overfitting (as viewed by decision surfaces) of decision trees with no limits on depth and leaf size</li>\n",
    "<li>Searching for optimal parameters of a decision tree</li>\n",
    "<li>Using decision tree output to assess feature importance</li>\n",
    "</ul>\n",
    "\n",
    "</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###Build a Tree on Simulated Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import scipy as sp\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "#These all need to be installed to both run and visualize a tree\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.tree import export_graphviz\n",
    "from IPython.display import Image\n",
    "import StringIO, pydot\n",
    "%matplotlib inline\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p>Here I generate some random data $X=<X1,X2>$ with $Y \\in [0,1]$\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x108c7c590>]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAEACAYAAAC08h1NAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x108bcca50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "'''\n",
    "Random data example \n",
    "'''\n",
    "n=200\n",
    "\n",
    "\n",
    "#First Generate the data\n",
    "\n",
    "m=[[0.25,0.75],[0.75,0.75],[0.75,0.25],[0.25,0.25]]\n",
    "s=.1\n",
    "X1=pd.DataFrame(np.random.multivariate_normal([m[0][0],m[0][1]],[[s,0],[0,s]],n), columns=['x1','x2'])\n",
    "X1['y']=np.zeros(n)\n",
    "X2=pd.DataFrame(np.random.multivariate_normal([m[1][0],m[1][1]],[[s,0],[0,s]],n), columns=['x1','x2'])\n",
    "X2['y']=np.zeros(n)\n",
    "X3=pd.DataFrame(np.random.multivariate_normal([m[2][0],m[2][1]],[[s,0],[0,s]],n), columns=['x1','x2'])\n",
    "X3['y']=np.ones(n)\n",
    "X4=pd.DataFrame(np.random.multivariate_normal([m[3][0],m[3][1]],[[s,0],[0,s]],n), columns=['x1','x2'])\n",
    "X4['y']=np.zeros(n)\n",
    "\n",
    "dat=X1.append(X2,ignore_index=True)\n",
    "dat=dat.append(X3,ignore_index=True)\n",
    "dat=dat.append(X4,ignore_index=True)\n",
    "\n",
    "plot(dat['x1'][(dat['y']==1)],dat['x2'][(dat['y']==1)],'r.')\n",
    "plot(dat['x1'][(dat['y']==0)],dat['x2'][(dat['y']==0)],'b.')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p>\n",
    "\n",
    "</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p>\n",
    "Now lets build a tree on our sample data.<br><br>\n",
    "\n",
    "Documentation can be found here:\n",
    "http://scikit-learn.org/stable/modules/generated/sklearn.tree.DecisionTreeClassifier.html#sklearn.tree.DecisionTreeClassifier\n",
    "<br>\n",
    "Also, here is another useful example:\n",
    "http://nbviewer.ipython.org/github/sujitpal/statlearning-notebooks/blob/master/src/chapter8.ipynb\n",
    "\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#We want Entropy as our splitting criteria\n",
    "clf = DecisionTreeClassifier(criterion='entropy',min_samples_leaf=1,max_depth=3)\n",
    "clf = clf.fit(dat.drop('y',1), dat.y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p>Building a tree is pretty easy. Visualizing the tree takes a little extra work and a few extra packages.<br><br>\n",
    "To plot it is necessary to install Pydot as well as graphviz: http://www.graphviz.org/\n",
    "If you have homebrew installed, run: brew install graphviz in the command line.\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABF0AAAGlCAYAAAAoHomBAABAAElEQVR4AeydCbhV0/vH3yZKIjKE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\nrkJggg==\n",
      "text/plain": [
       "<IPython.core.display.Image at 0x1090b8d10>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dot_data = StringIO.StringIO()\n",
    "export_graphviz(clf, out_file=dot_data)\n",
    "graph = pydot.graph_from_dot_data(dot_data.getvalue())\n",
    "Image(graph.create_png())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p>The decision surface can be derived by using the predict function and not having to parse the DT rules.<br><br>\n",
    "This code was adapted from http://scikit-learn.org/0.11/auto_examples/tree/plot_iris.html\n",
    "\n",
    "\n",
    "\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x105cc60d0>]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD7CAYAAACRxdTpAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQIRwFwgEggrk/wf6xoWyYwympwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x109b5f390>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X=dat.drop('y',1)\n",
    "y=dat.y\n",
    "\n",
    "plot_step = 0.02\n",
    "\n",
    "x_min, x_max = X['x1'].min(), X['x1'].max()\n",
    "y_min, y_max = X['x2'].min(), X['x2'].max()\n",
    "xx, yy = np.meshgrid(np.arange(x_min, x_max, plot_step),np.arange(y_min, y_max, plot_step))\n",
    "\n",
    "Z = clf.predict(np.c_[xx.ravel(), yy.ravel()])\n",
    "Z = Z.reshape(xx.shape)\n",
    "\n",
    "\n",
    "plt.subplot(111)\n",
    "\n",
    "cs = plt.contourf(xx, yy, Z, cmap=plt.cm.cool)\n",
    "plt.plot(dat['x1'][(dat['y']==1)],dat['x2'][(dat['y']==1)],'r.')\n",
    "plt.plot(dat['x1'][(dat['y']==0)],dat['x2'][(dat['y']==0)],'b.')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p>The above is a pretty simple tree, which we can see in the above plot.<br><br>\n",
    "\n",
    "As we allow for more complex trees, visualing the tree becomes very cumbersome, but we can still look at the decision surface. \n",
    "\n",
    "\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaAAAAEZCAYAAADR8/HkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x108da9e90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def treeComplex(X,y,leaf,dep,i):\n",
    "    clf = DecisionTreeClassifier(criterion='entropy',min_samples_leaf=leaf,max_depth=dep)\n",
    "    clf = clf.fit(X,y)\n",
    "    plot_step = 0.02\n",
    "    x_min, x_max = X['x1'].min(), X['x1'].max()\n",
    "    y_min, y_max = X['x2'].min(), X['x2'].max()\n",
    "    xx, yy = np.meshgrid(np.arange(x_min, x_max, plot_step),np.arange(y_min, y_max, plot_step))\n",
    "    Z = clf.predict(np.c_[xx.ravel(), yy.ravel()])\n",
    "    Z = Z.reshape(xx.shape)\n",
    "\n",
    "    ax = fig.add_subplot(3,3,i)\n",
    "    cs = plt.contourf(xx, yy, Z, cmap=plt.cm.cool)\n",
    "    ax.axes.get_xaxis().set_visible(False)\n",
    "    ax.axes.get_yaxis().set_visible(False)\n",
    "    plt.title('Dep={},Leaf={}'.format(dep,leaf))\n",
    "              \n",
    "leafs=[100,10,1]\n",
    "deps=[3,10,50]\n",
    "\n",
    "i=1\n",
    "fig=plt.figure()\n",
    "\n",
    "for l in leafs:\n",
    "    for d in deps:\n",
    "        treeComplex(dat.drop('y',1),dat.y,l,d,i)\n",
    "        i+=1\n",
    "        \n",
    "fig.tight_layout()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p>As we let the trees grow without any restrictions, we notice that the decision surface gets very fragmented. This defies our intuition that nearest neighbors of a point should have the same class assignment. Since we know that this data was generated from a few simple normal mixtures, we should understand that these more complex surfaces are terribly over fit.\n",
    "<br><br>\n",
    "Beyond toy examples, let's see this on real data.\n",
    "</p>\n",
    "\n",
    "###Decision Tree on Real Data\n",
    "<p>We don't just want to build a tree, we want to optimize the configuration of the tree. We'll use AUC as a metric on our test set and vary the max_depth and min_leaf_size. <br><br>\n",
    "\n",
    "The underlying data has a very rare outcome (less than 5%) so we'll down sample the data first to reach a 50/50 split.\n",
    "\n",
    "\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn.metrics import confusion_matrix,roc_auc_score\n",
    "import course_utils as bd\n",
    "reload(bd)\n",
    "\n",
    "#Load data and downsample for a 50/50 split, then split into a train/test\n",
    "f='/Users/briand/Desktop/ds course/datasets/ads_dataset_cut.txt'\n",
    "\n",
    "train_split = 0.5\n",
    "tdat = pd.read_csv(f,header=0,sep='\\t')\n",
    "\n",
    "moddat = bd.downSample(tdat,'y_buy',9)\n",
    "#We know the dataset is sorted so we can just split by index\n",
    "train = moddat[:int(math.floor(moddat.shape[0]*train_split))]\n",
    "test = moddat[int(math.floor(moddat.shape[0]*train_split)):]\n",
    "\n",
    "def testTrees(X_train,y_train,X_test,y_test,dep,leaf,auc):\n",
    "    clf = DecisionTreeClassifier(criterion='entropy',min_samples_leaf=leaf,max_depth=dep)\n",
    "    clf = clf.fit(X_train,y_train)\n",
    "    if (auc==0):\n",
    "        cm = confusion_matrix(clf.predict(X_test),y_test)\n",
    "        return (cm[0][0]+cm[1][1])/float(sum(cm))\n",
    "    else:\n",
    "        return roc_auc_score(y_test,clf.predict_proba(X_test)[:,1])\n",
    "    \n",
    "\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x10b01b610>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY0AAAEZCAYAAABrUHmEAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEGKx/wPn65CAXE7PFwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10994c190>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "lab='y_buy'\n",
    "\n",
    "depths=[4,5,10,20]\n",
    "leaves=np.arange(1,101)\n",
    "\n",
    "#Run all of the options\n",
    "run=1\n",
    "if (run==1):\n",
    "    #Initialize dictionary of results\n",
    "    res=dict()\n",
    "    for d in depths:\n",
    "        res[d]=list()\n",
    "\n",
    "    #Now train and get results for each option\n",
    "    for d in depths:\n",
    "        for l in leaves:\n",
    "            res[d].append(testTrees(train.drop(lab,1),train[lab],test.drop(lab,1),test[lab],d,l,1))\n",
    "\n",
    "\n",
    "#Now plot            \n",
    "fig = plt.figure()\n",
    "ax=fig.add_subplot(111)\n",
    "plt.plot(leaves,res[depths[0]],'b-',label='Depth={}'.format(depths[0]))\n",
    "plt.plot(leaves,res[depths[1]],'r-',label='Depth={}'.format(depths[1]))\n",
    "plt.plot(leaves,res[depths[2]],'y-',label='Depth={}'.format(depths[2]))\n",
    "plt.plot(leaves,res[depths[3]],'g-',label='Depth={}'.format(depths[3]))\n",
    "plt.legend(loc=4)\n",
    "ax.set_xlabel('Min Leaf Size')\n",
    "ax.set_ylabel('Test Set AUC')\n",
    "plt.title('Holdout AUC by Hyperparameters')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p>The above plot shows how holdout AUC varies by parameters. As we generally expect with most hyper-parameter searches, there is a \"sweet spot\" between letting the model be flexible and over fitting. Here we see that as long as we restrict the min_leaf_size, the max_depth doesn't affect the optimal choice too much.<br><br>\n",
    "\n",
    "We might want to peak at the tree to determine which features it thinks are important. The DT class objects actually have an attribute that gives each feature a score for how important it is in building the tree. \n",
    "\n",
    "\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x1067e1f50>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYQAAAFjCAYAAAA9/HzpAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x108b829d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "'''\n",
    "The tree gives us a useful tool for analying feature importance\n",
    "'''\n",
    "fig, ax = plt.subplots()\n",
    "width=0.35\n",
    "#ax.bar(train.drop(lab,1).columns.values, clf.feature_importances_, width, color='r')\n",
    "ax.bar(np.arange(13), clf.feature_importances_, width, color='r')\n",
    "ax.set_xticks(np.arange(len(clf.feature_importances_)))\n",
    "ax.set_xticklabels(train.drop(lab,1).columns.values,rotation=90)\n",
    "plt.title('Feature Importance from DT')\n",
    "ax.set_ylabel('Normalized Gini Importance')\n",
    "\n"
   ]
  }
 ],
 "metadata": {},
 "nbformat": 4,
 "nbformat_minor": 0
}