Final_Levato-Davis-Perkins.ipynb

gistfile1.txt

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Forest Cover Project \n",
    "### Ashley L, Kevin D, Eric P"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This notebook contains the final project work building a classification model for the Forest Cover Kaggle competition. The main part of the notebooks walks through some data visualization and methods that resulted in our highest accruacy model. An appendix is included at the end of the notebook which contains additional work and methods that were not part of the final model, but showcase other work that went into this final project. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.cross_validation import train_test_split\n",
    "from sklearn import preprocessing\n",
    "from sklearn import metrics\n",
    "from sklearn.ensemble import ExtraTreesClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.metrics import confusion_matrix\n",
    "import matplotlib.pyplot as plt\n",
    "import re\n",
    "from collections import defaultdict\n",
    "from matplotlib.colors import ListedColormap\n",
    "%matplotlib inline\n",
    "pd.set_option('display.max_columns', None)\n",
    "pd.set_option('display.float_format', lambda x: '%.2f' % x)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data importing and organization"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "Below we split the data into training and development sets to use in model testing. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "train_data= pd.read_csv('.//data//train.csv')\n",
    "train_data.head()\n",
    "test = pd.read_csv('.//data//test.csv')\n",
    "train_labels = train_data[\"Cover_Type\"].values\n",
    "\n",
    "\n",
    "train, dev = train_test_split(train_data, test_size = 0.2)\n",
    "columns = [i for i in train.columns if i not in ('Id', 'Cover_Type') ]\n",
    "\n",
    "# Test labels and features\n",
    "labels = train[\"Cover_Type\"].values\n",
    "features = train[list(columns)].values\n",
    "\n",
    "# Dev labels and features\n",
    "dev_features = dev[list(columns)].values\n",
    "dev_labels = dev[\"Cover_Type\"].values\n",
    "\n",
    "test = pd.read_csv('.//data//test.csv')\n",
    "\n",
    "test_features = test[list(columns)].values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Below we split the training data into the cover types in order to use it to color some visualizations below"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#split the data by the cover type\n",
    "#so we can create some visuals to look at the data\n",
    "train1=train[labels==1]\n",
    "train2=train[labels==2]\n",
    "train3=train[labels==3]\n",
    "train4=train[labels==4]\n",
    "train5=train[labels==5]\n",
    "train6=train[labels==6]\n",
    "train7=train[labels==7]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Initial Visualizations"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In order to get a better sense of what our data looked like and which variables might be of interested, we created some data visualizations. We started by plotting the historgrams of the variables and color coding the data based on the cover type from our trianing data. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#create a function to color the variable histrogram (for the overall data)\n",
    "#by the color of the label\n",
    "#this will give us some intution on which variables will be useful predictors of class\n",
    "\n",
    "def colorHist(dfs, col, ax, b=10, dropBin1=False):\n",
    "    \"\"\"\n",
    "    dfs: a list of the data frames by cover type\n",
    "    col: which column we want to plot\n",
    "    ax: passing the axes for plotting\n",
    "    b: number of bins, default is 10\n",
    "    dropBin1: this is if we only want to see the distribution of true elements \n",
    "              This is useful for visualizing soil and wilderness binary variables \n",
    "              where we only want to see the cover types for =1.\n",
    "              Default = False because in most cases we want all of the bins\n",
    "    \"\"\"\n",
    "    \n",
    "    if not isinstance(dfs,list):\n",
    "        return\n",
    "    colors = ['red','blue','lime','yellow','deeppink','purple','aqua']\n",
    "    counts, divs = np.histogram(train_data[col], bins=b)\n",
    "    if dropBin1:\n",
    "        b = b-1\n",
    "        divs = divs[1:]\n",
    "    bottoms = np.zeros(b)\n",
    "    for i in range(len(dfs)):\n",
    "        df = dfs[i]\n",
    "        sub_counts, divs = np.histogram(df[col], bins=b, range=(divs[0],divs[-1]))\n",
    "        ax.bar(left=divs[:-1], height=sub_counts, width=divs[1]-divs[0], bottom=bottoms, color=colors[i])\n",
    "        bottoms += sub_counts"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We visualized all of the variables during our exploration and model building which helped identify where to focus our efforts. A sample of just a few visuals are provided below. In order to view other variables add the column name to the cols list. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Batman\\Anaconda2\\lib\\site-packages\\matplotlib\\figure.py:1718: UserWarning: This figure includes Axes that are not compatible with tight_layout, so its results might be incorrect.\n",
      "  warnings.warn(\"This figure includes Axes that are not \"\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsgAAAKACAYAAACBoI53AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nWiyQJUmSpBYLZEmSJKnl/wMf7uE5DvReIgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x14eb73b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot the continuous variables\n",
    "\n",
    "cols = ['Elevation', 'Aspect', 'Wilderness_Area4','Soil_Type29']\n",
    "fig, axesarr = plt.subplots(2,2)\n",
    "fig.set_figheight(9)\n",
    "fig.set_figwidth(10)\n",
    "fig.set_tight_layout('tight')\n",
    "axes = np.hstack(axesarr)\n",
    "for ax, col in zip(axes, cols):\n",
    "    colorHist([train1, train2, train3, train4, train5, train6, train7], col=col, ax=ax, b=10)\n",
    "    ax.set_title(col)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A quick glance from the above we can see that the elevation is important in determining the type. The below plots breakdown the histogram of Elevation into different size bins, to help identify possible break points for binning the variable. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsgAAAKACAYAAACBoI53AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQZYkSZL6mCBLkiRJff4/aaXAeQsB8qkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x14eb7390>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "bins =[5, 10, 20, 30]\n",
    "\n",
    "fig, axesarr = plt.subplots(2,2)\n",
    "fig.set_figheight(9)\n",
    "fig.set_figwidth(10)\n",
    "fig.set_tight_layout('tight')\n",
    "axes = np.hstack(axesarr)\n",
    "for ax, bins in zip(axes, bins):\n",
    "    colorHist([train1, train2, train3, train4, train5, train6, train7], col = 'Elevation', ax=ax, b=bins)\n",
    "    ax.set_title(bins)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Other things we noted from some of the visuals not depicted here included a quick glance at the wilderness areas shows that area 3 is likely not going to be a great predictor because it contains almost all of the cover types. However, 1 and 4 may be able to help distinguish between some cover types. The plots of the soil variables show us that some soil types are very rare, but highly predictive when they occur (Soil Type 12, Soil Type 35), while other soil types are very common and roughly evenly distributed between all of the cover types (Soil Type 7, Soil Type 8). These variables may simply be non-informative and better left out of the model. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Pairwise Data Visualizations"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Additionally, the funciton below was developed to plot the pairwise variables and color based on cover type. This may be able to provide insights into pairs of variables that could help distinguish two or more cover types. The best use for this is after we have run some of the model and find out which cover types are being misclassified. This visualization can then be run to help find which variables to focus on for futher modeling and analysis. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def pairwisePlot(dfs,columns,save_fig_path=None):\n",
    "    colors = ['red','blue','lime','yellow','deeppink','purple','aqua']\n",
    "    if not all([isinstance(df, pd.core.frame.DataFrame) for df in dfs]):\n",
    "        raise TypeError('all elements of dfs must be a pandas DataFrames')\n",
    "    if not isinstance(columns, (list, tuple, np.ndarray)):\n",
    "        raise TypeError('columns must be a list, tuple, or numpy array')\n",
    "    dim = len(columns)\n",
    "    fig, axes = plt.subplots(dim, dim, figsize=(3*dim,3*dim))\n",
    "    fig.subplots_adjust(wspace=0.02, hspace=0.02)\n",
    "    for row in range(dim):\n",
    "        for col in range(dim):\n",
    "            ax = axes[row,col]\n",
    "            if row == col:\n",
    "                s = columns[row]\n",
    "                s0 = '\\n'.join([s[x*16:(x+1)*16] for x in range(1+len(s)/16)])\n",
    "                ax.text(x=0,y=0,s=s0,horizontalalignment='center')\n",
    "                ax.set_xlim(-1,1)\n",
    "                ax.set_ylim(-1,1)\n",
    "                ax.set_xticks([])\n",
    "                ax.set_yticks([])\n",
    "            else:\n",
    "                for i,df in enumerate(dfs):\n",
    "                    ax.scatter(df[columns[col]], df[columns[row]], color=colors[i],s=10, linewidths=0.3, alpha=.2)\n",
    "                ax.set_xticklabels([])\n",
    "                ax.set_yticklabels([])\n",
    "    if not save_fig_path is None:\n",
    "        fig.savefig(save_fig_path)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the pairwisePlot function, we and choose what cover types to plot and which variables to plot pairwise. Other analysis indicates the differences between cover type 1 and 2 may help increase accuracy. Below is an eample of the pairwise plots of a few variables for cover types 1 and 2."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAqwAAAKsCAYAAAApwu8wAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nBxIDt45jqAQpAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1bcba8d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dfs =[train1, train2]\n",
    "col = ['Elevation', 'Aspect', 'Slope', 'Wilderness_Area2']\n",
    "pairwisePlot(dfs[0:2],col,'./out.jpg')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the graphs above elevation seems to be a strong distinguishing variable. Additionally, Wilderness area 2 and several of the continuous vairables seem to split apart the two cover types. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Best Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Model Development\n",
    "One fact that emerged from our initial submissions was that the accuracy from predicting the development set was much higher than the accuracy for predicting the test set provided by kaggle. Further, the accuracy on the testing set seemed to be correlated to the accuracy of predicting classes one and two on the development set. This suggested that the uniform distribution of the classes in the training set was not reflecting in the test set, and trying to reflect the proper distribution would include model results. Optimizing the ratio of classes one and two to the other classes provided the biggest improvement in our score. The second big improvement was to create two models, one using the oversampling technique and the other with uniform distribution of classes. Using the oversampling model to make all predictions for classes one and two, and the uniform model to predict classes 3 through 6, significantly increased accuracy. Marginal improvements came from chosing interaction terms that had a positive effect on disambiguating classes one and two and removing non-informative soil features. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ExtraTreesClassifier(bootstrap=False, class_weight=None, criterion='gini',\n",
       "           max_depth=None, max_features='auto', max_leaf_nodes=None,\n",
       "           min_samples_leaf=1, min_samples_split=2,\n",
       "           min_weight_fraction_leaf=0.0, n_estimators=250, n_jobs=1,\n",
       "           oob_score=False, random_state=0, verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Add interaction and square terms\n",
    "\n",
    "train_data['Elevation_x_Horizontal_Distance_To_Roadways']= (\n",
    "    train_data['Elevation'] * train_data['Horizontal_Distance_To_Roadways'])\n",
    "train_data['Elevation_x_Horizontal_Distance_To_Fire_Points'] = (\n",
    "    train_data['Elevation']) * train_data['Horizontal_Distance_To_Fire_Points']\n",
    "train_data['Slope_x_Horizontal_Distance_To_Hydrology'] = (\n",
    "    train_data['Slope'] * train_data['Horizontal_Distance_To_Hydrology'])\n",
    "train_data['Horizontal_Distance_To_Fire_Points_squared'] =(\n",
    "    train_data['Horizontal_Distance_To_Fire_Points'] * train_data['Horizontal_Distance_To_Fire_Points'])\n",
    "train_data['Horizontal_Distance_To_Roadways_squared'] = (\n",
    "    train_data['Horizontal_Distance_To_Roadways'] * train_data['Horizontal_Distance_To_Roadways'])\n",
    "test['Elevation_x_Horizontal_Distance_To_Roadways']= (\n",
    "    test['Elevation'] * test['Horizontal_Distance_To_Roadways'])\n",
    "test['Elevation_x_Horizontal_Distance_To_Fire_Points'] = (\n",
    "    test['Elevation']) * test['Horizontal_Distance_To_Fire_Points']\n",
    "test['Slope_x_Horizontal_Distance_To_Hydrology'] = (\n",
    "    test['Slope'] * test['Horizontal_Distance_To_Hydrology'])\n",
    "test['Horizontal_Distance_To_Fire_Points_squared'] =(\n",
    "    test['Horizontal_Distance_To_Fire_Points'] * test['Horizontal_Distance_To_Fire_Points'])\n",
    "test['Horizontal_Distance_To_Roadways_squared'] = (\n",
    "    test['Horizontal_Distance_To_Roadways'] * test['Horizontal_Distance_To_Roadways'])\n",
    "\n",
    "# Subset the columns of interest, removing soil classes 7 and 8 \n",
    "cols = (\n",
    "    list(train_data.columns[1:21]) + train_data.columns[23:55].tolist() +\n",
    "    train_data.columns[56:].tolist())\n",
    "\n",
    "# Function for chosing which model prediction to use\n",
    "def voting(row):\n",
    "    if row['wt_preds'] == 2 or row['wt_preds'] == 1:\n",
    "        return row['wt_preds']\n",
    "    else:\n",
    "        return row['unwt_preds']\n",
    "\n",
    "# Extract features and labels for un-weighted model\n",
    "unwt_train_data = train_data.copy()\n",
    "unwt_train_features = unwt_train_data[cols].values\n",
    "unwt_train_labels = unwt_train_data['Cover_Type']\n",
    "\n",
    "# Create weighted training set which results in classes 1 and 2 being oversampled\n",
    "train1=train_data[train_data['Cover_Type']==1]\n",
    "train2=train_data[train_data['Cover_Type']==2]\n",
    "train3=train_data[train_data['Cover_Type']==3]\n",
    "train4=train_data[train_data['Cover_Type']==4]\n",
    "train5=train_data[train_data['Cover_Type']==5]\n",
    "train6=train_data[train_data['Cover_Type']==6]\n",
    "train7=train_data[train_data['Cover_Type']==7]\n",
    "mini_train3 = train3.sample(frac = 0.15)\n",
    "mini_train4 = train4.sample(frac = 0.15)\n",
    "mini_train5 = train5.sample(frac = 0.15)\n",
    "mini_train6 = train6.sample(frac = 0.15)\n",
    "mini_train7 = train7.sample(frac = 0.15)\n",
    "weighted_train = train1.append(train2).append(\n",
    "    mini_train3).append(mini_train4).append(\n",
    "    mini_train5).append(mini_train6).append(mini_train7)\n",
    "\n",
    "# Pull out the weighted training labels and order them correctly\n",
    "weighted_train = weighted_train.sort_values('Id')\n",
    "wt_train_features = weighted_train[cols].values\n",
    "weighted_train_labels = pd.read_csv('.\\\\data\\\\train.csv')\n",
    "weighted_train_labels = weighted_train_labels.loc[weighted_train_labels['Id'].isin(weighted_train['Id'])]\n",
    "weighted_train_labels = weighted_train_labels['Cover_Type'].values\n",
    "\n",
    "# Fit ERT models\n",
    "unwt_forest = ExtraTreesClassifier(n_estimators=250, random_state=0)\n",
    "unwt_forest.fit(unwt_train_features, unwt_train_labels)\n",
    "\n",
    "wt_forest = ExtraTreesClassifier(n_estimators=250, random_state=0)\n",
    "wt_forest.fit(wt_train_features, weighted_train_labels)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Predict and Output for Submission"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Predict test data\n",
    "test_features = test[cols].values\n",
    "unwt_preds = unwt_forest.predict(test_features)\n",
    "wt_pred = wt_forest.predict(test_features)\n",
    "\n",
    "# Output predictions to dataframe\n",
    "preds = pd.DataFrame(test['Id'])\n",
    "preds['unwt_preds'] = unwt_preds\n",
    "preds['wt_preds'] = wt_pred\n",
    "\n",
    "# Chose final predictions\n",
    "preds['pred'] = preds.apply(voting, axis=1)\n",
    "\n",
    "#Output to csv\n",
    "output = pd.DataFrame(test['Id'])\n",
    "output['Cover_Type'] = preds['pred']\n",
    "output.reset_index(inplace=True)\n",
    "del output['index']\n",
    "output.to_csv('.\\\\data\\\\submissions\\\\sub2_04232016.csv', index =False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test Set Accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABRUAAADZCAYAAAC+atcbAAAAAXNSR0IArs4c6QAAAARnQU1BAACx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\nuGaMTCd3AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import Image\n",
    "Image(filename=r\"data\\submissions\\sub2_04232016.PNG\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Appendix of Additional Work"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A lot of other work went into finding our best model, some things that did not work so well or were not needed for the best model are included here for reference. This appendix covers some model testing, misclassification analysis, soil analysis and data cleaning. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Model Building"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Select columns\n",
    "cols = [\n",
    "    [i for i in train.columns if i not in ('Id', 'Cover_Type', 'index') ],\n",
    "    [i for i in train.columns if i not in ('Id', 'Cover_Type', 'index', 'Wilderness_Area2', 'Wilderness_Area3',\n",
    "                                          'Soil_Type7', 'Soil_Type8', 'Soil_Type15') ]]\n",
    "#Create vectors of features and labels for the training and dev data\n",
    "\n",
    "labels = train[\"Cover_Type\"].values\n",
    "features = [\n",
    "    train[list(cols[0])].values, train[list(cols[1])].values]\n",
    "dev_labels = dev[\"Cover_Type\"].values\n",
    "dev_features = [\n",
    "    dev[list(cols[0])].values, dev[list(cols[0])].values]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From other analysis the ExtraTreesClassifier performs well and also normalizes the data. Normalization provided in the Appendix for running other models. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "One option we explored was weighting features one and two more heavily in the model. This produced a small improvemnt, but was much less effective than oversampling classes one and two. Weighting, combined with oversampling, did not improve accuracy. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Going to try increasing the weights for classes 1 and 2 to see if \n",
    "# we can improve performance on those classes specifically\n",
    "weights = np.array([1, 2, 4, 8, 10, 100, 1000, 10000, 100000, 1000000])\n",
    "scores = np.zeros(weights.shape[0])\n",
    "plt_counter = 1\n",
    "fig = plt.figure(figsize=(12, 20))\n",
    "fig.suptitle('Confusion Matrix for Different Weights of Class 1 and 2', fontsize=14, fontweight='bold')\n",
    "for w in weights:\n",
    "    lab_wght = {1:w, 2:w}\n",
    "    forest = ExtraTreesClassifier(n_estimators=250, random_state=0,class_weight = lab_wght)\n",
    "    forest.fit(features[0], labels)\n",
    "    scores[np.where(weights==w)[0][0]] = forest.score(dev_features[0], dev_labels)\n",
    "    preds = forest.predict(dev_features[0])\n",
    "    #Get the confusion matrix\n",
    "    cm = confusion_matrix(dev_labels, preds)\n",
    "    cm_normalized = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "    ax = fig.add_subplot(4,3,plt_counter)\n",
    "    ax.set_title(\"Class 1 and 2 Feature Weights: {}\\nAccuracy: {}\".format(\n",
    "            w, forest.score(dev_features[0], dev_labels)))\n",
    "    ax.imshow(cm_normalized, interpolation='nearest', cmap=plt.cm.Blues)\n",
    "    for (j,i), label in np.ndenumerate(cm_normalized):\n",
    "        if i == j:\n",
    "            col = 'w'\n",
    "        else:\n",
    "            col = 'black'\n",
    "        ax.text(i,j, round(label,3), ha='center', va='center', color = col)\n",
    "    plt_counter += 1\n",
    "plt.show()\n",
    "print scores"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Trying different balancing schemes for the weighting also increased accuracy marginally, but not after oversampling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "balance = np.array([\"auto\", \"subsample\"])\n",
    "scores = np.zeros(balance.shape[0])\n",
    "fig = plt.figure(figsize = (8, 8))\n",
    "plt_counter = 1\n",
    "for b in balance:\n",
    "    forest = ExtraTreesClassifier(n_estimators=250, random_state=0,class_weight = b)\n",
    "    forest.fit(features[0], labels)\n",
    "    scores[np.where(balance==b)[0][0]] = forest.score(dev_features[0], dev_labels)\n",
    "    preds = forest.predict(dev_features[0])\n",
    "    #Get the confusion matrix\n",
    "    cm = confusion_matrix(dev_labels, preds)\n",
    "    cm_normalized = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "    ax = fig.add_subplot(1,2,plt_counter)\n",
    "    ax.set_title(\"Feature Weighting Method: {}\\nAccuracy: {}\".format(\n",
    "            b, forest.score(dev_features[0], dev_labels)))\n",
    "    ax.imshow(cm_normalized, interpolation='nearest', cmap=plt.cm.Blues)\n",
    "    for (j,i), label in np.ndenumerate(cm_normalized):\n",
    "        if i == j:\n",
    "            col = 'w'\n",
    "        else:\n",
    "            col = 'black'\n",
    "        ax.text(i,j, round(label,3), ha='center', va='center', color = col)\n",
    "    plt_counter += 1\n",
    "plt.show()\n",
    "print scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "feature_dict = dict(zip(cols[0], forest.feature_importances_))\n",
    "cols = ([i for i in cols[0] if feature_dict[i] > 0])\n",
    "\n",
    "forest2 = ExtraTreesClassifier(n_estimators=250,\n",
    "                              random_state=0)\n",
    "labels = train[\"Cover_Type\"].values\n",
    "features = train[cols].values\n",
    "dev_lables = dev[\"Cover_Type\"].values\n",
    "dev_features = dev[cols].values\n",
    "\n",
    "forest2.fit(features, labels)\n",
    "print forest2.score(dev_features, dev_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Misclassification Analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From our best model at the mid-term of the project we conducted some missclasification analysis. This can help us figure out how to re-weight the model or possible areas where we need to funnel the results into secondary models. In this case, we see that most of the results are from mistakes between cover types 1/2 and 3/6.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from collections import defaultdict\n",
    "    \n",
    "def misClass(forest, dev_features, dev_labels):\n",
    "    pred = forest.predict(dev_features)\n",
    "    pred_probs = forest.predict_proba(dev_features)\n",
    "    pair_dict = defaultdict(int)\n",
    "    bool = np.array(pred!=dev_labels)\n",
    "    for p,dl in zip(pred_probs[bool],dev_labels[bool]) :\n",
    "        pair_dict['Prediced %d -> Actual %d' %(np.argmax(p)+1, dl)] += 1\n",
    "    print pair_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "misClass(forest2, dev_features, dev_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "From the results above we see that the most mistakes are being made between 1 and 2, followed by 3 and 6. If we run some of the pairwise visualizations we can pick our variables that may help us distinguish these classes. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "dfs =[train1, train2]\n",
    "col = ['Elevation', 'Aspect', 'Slope', 'Horizontal_Distance_To_Hydrology', 'Vertical_Distance_To_Hydrology', \n",
    "        'Horizontal_Distance_To_Roadways', 'Hillshade_9am', 'Hillshade_Noon', 'Hillshade_3pm', 'Horizontal_Distance_To_Fire_Points', \n",
    "        'Wilderness_Area1', 'Wilderness_Area2', 'Wilderness_Area3', 'Wilderness_Area4']\n",
    "pairwisePlot(dfs[0:2],col,'./out.jpg')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the output above we can see elevation seems to be a strong distinguishing variable along with Wilderness area 2 and several of the continuous vairables seem to split apart the two cover types. Note that in many cases, chosing an interaction term that marginally improves classification in two groups resulted in decreased accuracy in other classes. \n",
    "\n",
    "We conducted similar analysis for cover types 3 and 6, the wilderness areas are not good distinguishing features. However, some of the continuous variables have less overlap than others. Additionally we could examine the soil types as possible distinguising charactersitics. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Feature Engineering"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To try and improve performance on classes one and two, we looked at using interaction terms that had a positive impact\n",
    "on disambiguating the two clasess."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "train.columns[1:56]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "ele_cross_cols = [u'Aspect', u'Slope',\n",
    "       u'Horizontal_Distance_To_Hydrology', u'Vertical_Distance_To_Hydrology',\n",
    "       u'Horizontal_Distance_To_Roadways', u'Hillshade_9am', u'Hillshade_Noon',\n",
    "       u'Hillshade_3pm', u'Horizontal_Distance_To_Fire_Points']\n",
    "\n",
    "for i in ele_cross_cols:\n",
    "    train['Elevation_x_' + str(i)] = train['Elevation'] * train[i]\n",
    "    dev['Elevation_x_' + str(i)] = dev['Elevation'] * dev[i]\n",
    "\n",
    "cols = [\n",
    "    train.columns[1:55],\n",
    "    list(train.columns[1:55]) + [train.columns[56]],\n",
    "    list(train.columns[1:55]) + [train.columns[57]],\n",
    "    list(train.columns[1:55]) + [train.columns[58]],\n",
    "    list(train.columns[1:55]) + [train.columns[59]],\n",
    "    list(train.columns[1:55]) + [train.columns[60]],\n",
    "    list(train.columns[1:55]) + [train.columns[61]],\n",
    "    list(train.columns[1:55]) + [train.columns[62]],\n",
    "    list(train.columns[1:55]) + [train.columns[63]],\n",
    "    list(train.columns[1:55]) + [train.columns[64]],\n",
    "    list(train.columns[1:55]) + train.columns[[60,64]].tolist()]\n",
    "#     list(train_data.columns[1:55]) + train_data.columns[[59,63, 72]].tolist(),]\n",
    "train_features = [train[cols[i]] for i in range(len(cols))]\n",
    "dev_features = [dev[cols[i]] for i in range(len(cols))]\n",
    "\n",
    "extra_cols = ['None'] + list(train.columns[56:]) + [\"Multiple\"]\n",
    "scores = np.zeros(len(dev_features))\n",
    "plt_counter = 1\n",
    "fig = plt.figure(figsize=(16, 18))\n",
    "fig.suptitle('Confusion Matrix for Different Interaction Terms', fontsize=14, fontweight='bold')\n",
    "for d in range(scores.shape[0]):\n",
    "    forest = ExtraTreesClassifier(n_estimators=250, random_state=0)\n",
    "    forest.fit(train_features[d], labels)\n",
    "    scores[d] = forest.score(dev_features[d], dev_labels)\n",
    "    preds = forest.predict(dev_features[d])\n",
    "    # Get confusion matrix\n",
    "    cm = confusion_matrix(dev_labels, preds)\n",
    "    cm_normalized = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "    ax = fig.add_subplot(4, 3, plt_counter)\n",
    "    ax.set_title(\"Column: {}\\nAccuracy: {}\".format(\n",
    "            extra_cols[d], forest.score(dev_features[d], dev_labels)))\n",
    "    ax.imshow(cm_normalized, interpolation='nearest', cmap=plt.cm.Blues)\n",
    "    for (j,i), label in np.ndenumerate(cm_normalized):\n",
    "        if i == j:\n",
    "            col = 'w'\n",
    "        else:\n",
    "            col = 'black'\n",
    "        ax.text(i,j, round(label,3), ha='center', va='center', color = col)\n",
    "    plt_counter += 1    \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that we by including two interaction terms, we improve class one and two performance by 3% in the dev set. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Model Breakouts for Missclassified Groups"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since cover types 1 and 2 are so hard to distinguish, one approach we tried was to take the origional model and if it predicts a 1 or a 2 then it runs a sub model, similarly with cover types 3 and 6. The function below will run the model. However, this model did not improve our overall best model, so we moved on to other ways to distinguish between these misclassfied cover types. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def run_multi_model(train_data, train_labels, test_data):\n",
    "    forest = ExtraTreesClassifier(n_estimators=250, random_state=0, criterion='entropy', bootstrap=False)\n",
    "    forest.fit(train_data, train_labels)\n",
    "    \n",
    "    forest12 = ExtraTreesClassifier(n_estimators=250, random_state=0, criterion='entropy', bootstrap=False)\n",
    "    forest12.fit(train_data[(train_labels==1)|(train_labels==2)], \n",
    "                 train_labels[(train_labels==1)|(train_labels==2)])\n",
    "    \n",
    "    forest36 = ExtraTreesClassifier(n_estimators=250, random_state=0, criterion='entropy', bootstrap=False)\n",
    "    forest36.fit(train_data[(train_labels==3)|(train_labels==6)], \n",
    "                 train_labels[(train_labels==3)|(train_labels==6)])\n",
    "    \n",
    "    prediction = forest.predict(test_data)\n",
    "    for index,p in enumerate(prediction):\n",
    "        if p==1 or p==2:\n",
    "            prediction[index]=forest12.predict(test_data[index])\n",
    "        if p==3 or p==6:\n",
    "            prediction[index]=forest36.predict(test_data[index])\n",
    "    return prediction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Soil Analysis and Feature Selection part 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Reviewing the various soil type descriptions, we saw that various distinct soil types were very similar to one another in certain characteristics (e.g. stoniness, complex, etc.). As a result, we attempted to capture this similarity/dissimilarity by creating dummy variables for features of the soil. Below is code to mark all \"stony\" soil types as such. All attempted methods of marking the soil in this way failed to increase accuracy."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def mark_stony(df):\n",
    "    #1.2.6.9.12.18.24.25-34.36-40\n",
    "    if (df[\"Soil_Type1\"] == 1 or df[\"Soil_Type2\"] == 1 or df[\"Soil_Type6\"] == 1 or df[\"Soil_Type9\"] == 1): return 1\n",
    "    if (df[\"Soil_Type12\"] == 1 or df[\"Soil_Type18\"] == 1 or df[\"Soil_Type24\"] == 1 or df[\"Soil_Type25\"] == 1): return 1\n",
    "    if (df[\"Soil_Type26\"] == 1 or df[\"Soil_Type27\"] == 1 or df[\"Soil_Type28\"] == 1 or df[\"Soil_Type29\"] == 1): return 1\n",
    "    if (df[\"Soil_Type30\"] == 1 or df[\"Soil_Type31\"] == 1 or df[\"Soil_Type32\"] == 1 or df[\"Soil_Type33\"] == 1): return 1\n",
    "    if (df[\"Soil_Type34\"] == 1 or df[\"Soil_Type36\"] == 1 or df[\"Soil_Type37\"] == 1 or df[\"Soil_Type38\"] == 1): return 1\n",
    "    if (df[\"Soil_Type39\"] == 1 or df[\"Soil_Type40\"] == 1): return 1\n",
    "    return 0\n",
    "# train[\"Stony\"] = train.apply(mark_stony, axis=1)\n",
    "# dev[\"Stony\"] = dev.apply(mark_stony, axis=1)\n",
    "\n",
    "def bin_elevation_str(df):\n",
    "    return str(df[\"Elevation\"] / 400)\n",
    "\n",
    "def bin_elevation_num(df):\n",
    "    return df[\"Elevation\"] / 400\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following code investigates whether binning the elevation feature offers improvement in accuracy. I created three sets of train/dev data. The first does not include a binned elevation, the second bins numerically, and the third bins with strings. The code chunch two blocks down analyzes the results via confusion matrices. While there were consistent differences in accuracy for each result, there was no consistent winner among the six possible permutations (six because we also had one set with more features and one with fewer features for each of the three). Since binning elevation consistently was helpful, I chose to bin numerically as this was even in performance with string binning. This is in the Appendix since another form of binning with greater accuracy was ultimately chosen."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "train_data= pd.read_csv('.//data//train.csv')\n",
    "train_data.head()\n",
    "test = pd.read_csv('.//data//test.csv')\n",
    "train_labels = train_data[\"Cover_Type\"].values\n",
    "\n",
    "\n",
    "train, dev = train_test_split(train_data, test_size = 0.2)\n",
    "columns = [i for i in train.columns if i not in ('Id', 'Cover_Type') ]\n",
    "\n",
    "# Test labels and features\n",
    "labels = train[\"Cover_Type\"].values\n",
    "features = train[list(columns)].values\n",
    "\n",
    "# Dev labels and features\n",
    "dev_features = dev[list(columns)].values\n",
    "dev_labels = dev[\"Cover_Type\"].values\n",
    "\n",
    "test = pd.read_csv('.//data//test.csv')\n",
    "\n",
    "test_features = test[list(columns)].values\n",
    "\n",
    "# Set labels\n",
    "train_labels = train[\"Cover_Type\"].values\n",
    "dev_labels = dev[\"Cover_Type\"].values\n",
    "\n",
    "#Set features1\n",
    "columns1 = [\n",
    "    [i for i in train.columns if i not in ('Id', 'Cover_Type', 'index') ],\n",
    "    [i for i in train.columns if i not in ('Id', 'Cover_Type', 'index', 'Wilderness_Area1', 'Wilderness_Area3',\n",
    "                                          'Soil_Type7', 'Soil_Type8', 'Soil_Type15') ]]\n",
    "train_features = [\n",
    "    train[list(columns1[0])].values, train[list(columns1[1])].values]\n",
    "dev_features = [\n",
    "    dev[list(columns1[0])].values, dev[list(columns1[1])].values]\n",
    "\n",
    "#Set features2\n",
    "train[\"Elevation2\"] = train.apply(bin_elevation_str, axis=1)\n",
    "dev[\"Elevation2\"] = dev.apply(bin_elevation_str, axis=1)\n",
    "columns2 = [\n",
    "    [i for i in train.columns if i not in ('Id', 'Cover_Type', 'index') ],\n",
    "    [i for i in train.columns if i not in ('Id', 'Cover_Type', 'index', 'Wilderness_Area1', 'Wilderness_Area3',\n",
    "                                          'Soil_Type7', 'Soil_Type8', 'Soil_Type15') ]]\n",
    "train_features2 = [\n",
    "    train[list(columns2[0])].values, train[list(columns2[1])].values]\n",
    "dev_features2 = [\n",
    "    dev[list(columns2[0])].values, dev[list(columns2[1])].values]\n",
    "\n",
    "#Set features3\n",
    "train[\"Elevation2\"] = train.apply(bin_elevation_num, axis=1)\n",
    "dev[\"Elevation2\"] = dev.apply(bin_elevation_num, axis=1)\n",
    "columns3 = [\n",
    "    [i for i in train.columns if i not in ('Id', 'Cover_Type', 'index') ],\n",
    "    [i for i in train.columns if i not in ('Id', 'Cover_Type', 'index', 'Wilderness_Area1', 'Wilderness_Area3',\n",
    "                                          'Soil_Type7', 'Soil_Type8', 'Soil_Type15') ]]\n",
    "train_features3 = [\n",
    "    train[list(columns3[0])].values, train[list(columns3[1])].values]\n",
    "dev_features3 = [\n",
    "    dev[list(columns3[0])].values, dev[list(columns3[1])].values]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "tsets = [train_features, train_features2, train_features3]\n",
    "dsets = [dev_features, dev_features2, dev_features3]\n",
    "plt_counter = 1\n",
    "fig = plt.figure(figsize=(12, 20))\n",
    "fig.suptitle('Confusion Matrix for Different Training Sets', fontsize=14, fontweight='bold')\n",
    "lab_wght = {1:2, 2:1}\n",
    "for i in range(3):\n",
    "    for j in range(2):\n",
    "        forest = ExtraTreesClassifier(n_estimators=250, random_state=0,class_weight = lab_wght)\n",
    "        forest.fit(tsets[i][j], train_labels)\n",
    "        preds = forest.predict(dsets[i][j])\n",
    "        #Get the confusion matrix\n",
    "        cm = confusion_matrix(dev_labels, preds)\n",
    "        cm_normalized = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "        ax = fig.add_subplot(4,3,plt_counter)\n",
    "        ax.set_title(\"Training Set {}, number {}\\nAccuracy: {}\".format(\n",
    "                i+1, j+1, forest.score(dsets[i][j], dev_labels)))\n",
    "        ax.imshow(cm_normalized, interpolation='nearest', cmap=plt.cm.Blues)\n",
    "        for (n,m), label in np.ndenumerate(cm_normalized):\n",
    "            if m == n:\n",
    "                col = 'w'\n",
    "            else:\n",
    "                col = 'black'\n",
    "            ax.text(m,n, round(label,3), ha='center', va='center', color = col)\n",
    "        plt_counter += 1\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Forest generator used in earlier iterations of our design. The first argument determines whether the model will have more selective or inclusive featuers, and the second determines number of estimators for tree."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def run_forest(findex, ntrees):\n",
    "    #Generate data based on columns\n",
    "    weights = {1:2, 2:1}\n",
    "    train_data = train_features2[findex]\n",
    "    dev_data = dev_features2[findex]\n",
    "    #Build forest\n",
    "    forest = ExtraTreesClassifier(n_estimators=ntrees,random_state=0, class_weight=weights)\n",
    "    forest.fit(train_data, train_labels)\n",
    "    print str(ntrees) + \" trees with features[\" + str(findex) + \"] score:\", forest.score(dev_data, dev_labels)\n",
    "    return forest\n",
    "\n",
    "#Prints score for basic forest, and returns forest for analysis of columns\n",
    "forest = run_forest(1, 300)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Soil Analysis part 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We also pulled in some additional data on the soil types, the plan was to explore clusting the soil types with additional informaiton and using these clusters as new variables instead of the origional 40 soil types. However, other analysis proved showed a lot of improvements and we ran out of time to continue exploring this avenue. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df = pd.read_csv('./data/train.csv')\n",
    "#collapsing the soil data into one column \n",
    "df_melt = pd.melt(df, id_vars='Id', value_vars=[col for col in df.columns if 'Soil_Type' in col],\n",
    "                  value_name='Soil_Type_Bool', var_name='Soil_Type_Str')\n",
    "df_melt = df_melt[df_melt.Soil_Type_Bool==1]\n",
    "df_melt['Soil_Type']=df_melt['Soil_Type_Str'].apply(lambda x: int(x.replace('Soil_Type','')))\n",
    "df_melt = df_melt.sort_values('Id').reset_index(drop=True)\n",
    "df = df.drop([col for col in df.columns if 'Soil_Type' in col], axis=1)\n",
    "df = df.merge(df_melt[['Id','Soil_Type']],how='left',on='Id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#soil data/descriptions from US Forest Service (USFS)\n",
    "df_soils = pd.read_csv('.//soil_data.tsv', sep='\\t')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#merged with data\n",
    "df = df.merge(df_soils[['Soil_Type','Climatic_Zone','Geologic_Zone']], how='left', on='Soil_Type')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#output to csv for furture referece\n",
    "df.to_csv('./data/train_processed_soil.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data Normalization "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Noramlization of data was required initially for some of our early models. However, the extra trees model normalizes the data so we moved this section of the code to the appendix for reference. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "df_pre = train[[u'Elevation', u'Aspect', u'Slope',\n",
    "       u'Horizontal_Distance_To_Hydrology', u'Vertical_Distance_To_Hydrology',\n",
    "       u'Horizontal_Distance_To_Roadways', u'Hillshade_9am', u'Hillshade_Noon',\n",
    "       u'Hillshade_3pm', u'Horizontal_Distance_To_Fire_Points']]\n",
    "df_std = preprocessing.StandardScaler().fit_transform(df_pre)\n",
    "df = pd.DataFrame(df_std)\n",
    "df.columns = [u'Elevation', u'Aspect', u'Slope',\n",
    "       u'Horizontal_Distance_To_Hydrology', u'Vertical_Distance_To_Hydrology',\n",
    "       u'Horizontal_Distance_To_Roadways', u'Hillshade_9am', u'Hillshade_Noon',\n",
    "       u'Hillshade_3pm', u'Horizontal_Distance_To_Fire_Points']\n",
    "\n",
    "ts_pre = test[[u'Elevation', u'Aspect', u'Slope',\n",
    "       u'Horizontal_Distance_To_Hydrology', u'Vertical_Distance_To_Hydrology',\n",
    "       u'Horizontal_Distance_To_Roadways', u'Hillshade_9am', u'Hillshade_Noon',\n",
    "       u'Hillshade_3pm', u'Horizontal_Distance_To_Fire_Points']]\n",
    "ts_std = preprocessing.StandardScaler().fit_transform(ts_pre)\n",
    "ts = pd.DataFrame(ts_std)\n",
    "ts.columns = [u'Elevation', u'Aspect', u'Slope',\n",
    "       u'Horizontal_Distance_To_Hydrology', u'Vertical_Distance_To_Hydrology',\n",
    "       u'Horizontal_Distance_To_Roadways', u'Hillshade_9am', u'Hillshade_Noon',\n",
    "       u'Hillshade_3pm', u'Horizontal_Distance_To_Fire_Points']\n",
    "\n",
    "dev_pre = dev[[u'Elevation', u'Aspect', u'Slope',\n",
    "       u'Horizontal_Distance_To_Hydrology', u'Vertical_Distance_To_Hydrology',\n",
    "       u'Horizontal_Distance_To_Roadways', u'Hillshade_9am', u'Hillshade_Noon',\n",
    "       u'Hillshade_3pm', u'Horizontal_Distance_To_Fire_Points']]\n",
    "dev_std = preprocessing.StandardScaler().fit_transform(dev_pre)\n",
    "dev_df = pd.DataFrame(dev_std)\n",
    "dev_df.columns = [u'Elevation', u'Aspect', u'Slope',\n",
    "       u'Horizontal_Distance_To_Hydrology', u'Vertical_Distance_To_Hydrology',\n",
    "       u'Horizontal_Distance_To_Roadways', u'Hillshade_9am', u'Hillshade_Noon',\n",
    "       u'Hillshade_3pm', u'Horizontal_Distance_To_Fire_Points']\n",
    "orig = train[[u'Id', u'Wilderness_Area1', u'Wilderness_Area2', u'Wilderness_Area3',\n",
    "       u'Wilderness_Area4', u'Soil_Type1', u'Soil_Type2', u'Soil_Type3',\n",
    "       u'Soil_Type4', u'Soil_Type5', u'Soil_Type6', u'Soil_Type7',\n",
    "       u'Soil_Type8', u'Soil_Type9', u'Soil_Type10', u'Soil_Type11',\n",
    "       u'Soil_Type12', u'Soil_Type13', u'Soil_Type14', u'Soil_Type15',\n",
    "       u'Soil_Type16', u'Soil_Type17', u'Soil_Type18', u'Soil_Type19',\n",
    "       u'Soil_Type20', u'Soil_Type21', u'Soil_Type22', u'Soil_Type23',\n",
    "       u'Soil_Type24', u'Soil_Type25', u'Soil_Type26', u'Soil_Type27',\n",
    "       u'Soil_Type28', u'Soil_Type29', u'Soil_Type30', u'Soil_Type31',\n",
    "       u'Soil_Type32', u'Soil_Type33', u'Soil_Type34', u'Soil_Type35',\n",
    "       u'Soil_Type36', u'Soil_Type37', u'Soil_Type38', u'Soil_Type39',\n",
    "       u'Soil_Type40', u'Cover_Type']]\n",
    "\n",
    "orig.reset_index(inplace=True)\n",
    "# orig.head()\n",
    "df_std = pd.concat([df, orig], axis=1)\n",
    "df_std['Cover_Type'].shape\n",
    "\n",
    "o_dev = dev[[u'Id', u'Wilderness_Area1', u'Wilderness_Area2', u'Wilderness_Area3',\n",
    "       u'Wilderness_Area4', u'Soil_Type1', u'Soil_Type2', u'Soil_Type3',\n",
    "       u'Soil_Type4', u'Soil_Type5', u'Soil_Type6', u'Soil_Type7',\n",
    "       u'Soil_Type8', u'Soil_Type9', u'Soil_Type10', u'Soil_Type11',\n",
    "       u'Soil_Type12', u'Soil_Type13', u'Soil_Type14', u'Soil_Type15',\n",
    "       u'Soil_Type16', u'Soil_Type17', u'Soil_Type18', u'Soil_Type19',\n",
    "       u'Soil_Type20', u'Soil_Type21', u'Soil_Type22', u'Soil_Type23',\n",
    "       u'Soil_Type24', u'Soil_Type25', u'Soil_Type26', u'Soil_Type27',\n",
    "       u'Soil_Type28', u'Soil_Type29', u'Soil_Type30', u'Soil_Type31',\n",
    "       u'Soil_Type32', u'Soil_Type33', u'Soil_Type34', u'Soil_Type35',\n",
    "       u'Soil_Type36', u'Soil_Type37', u'Soil_Type38', u'Soil_Type39',\n",
    "       u'Soil_Type40', u'Cover_Type']]\n",
    "o_dev.reset_index(inplace=True)\n",
    "dev_std = pd.concat([dev_df, o_dev], axis=1)\n",
    "\n",
    "o_test = test[[u'Id', u'Wilderness_Area1', u'Wilderness_Area2', u'Wilderness_Area3',\n",
    "       u'Wilderness_Area4', u'Soil_Type1', u'Soil_Type2', u'Soil_Type3',\n",
    "       u'Soil_Type4', u'Soil_Type5', u'Soil_Type6', u'Soil_Type7',\n",
    "       u'Soil_Type8', u'Soil_Type9', u'Soil_Type10', u'Soil_Type11',\n",
    "       u'Soil_Type12', u'Soil_Type13', u'Soil_Type14', u'Soil_Type15',\n",
    "       u'Soil_Type16', u'Soil_Type17', u'Soil_Type18', u'Soil_Type19',\n",
    "       u'Soil_Type20', u'Soil_Type21', u'Soil_Type22', u'Soil_Type23',\n",
    "       u'Soil_Type24', u'Soil_Type25', u'Soil_Type26', u'Soil_Type27',\n",
    "       u'Soil_Type28', u'Soil_Type29', u'Soil_Type30', u'Soil_Type31',\n",
    "       u'Soil_Type32', u'Soil_Type33', u'Soil_Type34', u'Soil_Type35',\n",
    "       u'Soil_Type36', u'Soil_Type37', u'Soil_Type38', u'Soil_Type39',\n",
    "       u'Soil_Type40']]\n",
    "o_test.reset_index(inplace=True)\n",
    "test_std = pd.concat([ts, o_test], axis=1)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}