/code/preprocessing.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "accel00001 = pd.read_csv('/SPHERE-Challenge/data/raw_data/public_data/train/00001/acceleration.csv')\n",
    "target00001 = pd.read_csv('/SPHERE-Challenge/data/raw_data/public_data/train/00001/targets.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "def window_feat(t_start, t_end, df, columns):\n",
    "    \n",
    "    ''' Extract statistical features for a given window and sensor data frame df '''\n",
    "    \n",
    "    window_size = t_end - t_start\n",
    "    window_data = df[(df['t'] >= t_start) & (df['t'] < t_end)]\n",
    "    window_data = window_data[columns]\n",
    "    \n",
    "    feat = [] # empty features list\n",
    "    fnames = [] # empty feature names list\n",
    "\n",
    "    # quantiles\n",
    "    feat = feat + window_data.quantile(q = [0, 0.25, 0.50, 0.75, 1]).values.flatten().tolist()\n",
    "    fnames = ( fnames + \n",
    "              [ (column + '_W' + str(window_size) + '_0Q') for column in window_data.columns] + \n",
    "              [ (column + '_W' + str(window_size) + '_25Q') for column in window_data.columns] +\n",
    "              [ (column + '_W' + str(window_size) + '_50Q') for column in window_data.columns] +\n",
    "              [ (column + '_W' + str(window_size) + '_75Q') for column in window_data.columns] +\n",
    "              [ (column + '_W' + str(window_size) + '_100Q') for column in window_data.columns] )\n",
    "    \n",
    "    # mean\n",
    "    feat = feat + window_data.mean().tolist()\n",
    "    fnames = fnames + [ (column + '_W' + str(window_size) + '_mean') for column in window_data.columns]\n",
    "    \n",
    "    # std\n",
    "    feat = feat + window_data.std().tolist()\n",
    "    fnames = fnames + [ (column + '_W' + str(window_size) + '_std') for column in window_data.columns]\n",
    "    \n",
    "    return feat, fnames"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "feats = []\n",
    "\n",
    "for idx in target00001.index: # [0,1823]\n",
    "    \n",
    "    s = target00001.loc[idx,'start']\n",
    "    e = target00001.loc[idx,'end']\n",
    "    \n",
    "    # extract features\n",
    "    feat_W1, fnames_W1 = window_feat(t_start = e - 1, t_end = e, df = accel00001, columns = ['x', 'y', 'z'])\n",
    "    feat_W2, fnames_W2 = window_feat(t_start = e - 2, t_end = e, df = accel00001, columns = ['x', 'y', 'z'])\n",
    "    feat_W3, fnames_W3 = window_feat(t_start = e - 3, t_end = e, df = accel00001, columns = ['x', 'y', 'z'])\n",
    "    feat_W4, fnames_W4 = window_feat(t_start = e - 4, t_end = e, df = accel00001, columns = ['x', 'y', 'z'])\n",
    "    \n",
    "    # consolidate features\n",
    "    feat = feat_W1 + feat_W2 + feat_W3 + feat_W4\n",
    "    fnames = fnames_W1 + fnames_W2 + fnames_W3 + fnames_W4\n",
    "    \n",
    "    feats.append(feat)\n",
    "    \n",
    "feats = pd.DataFrame(feats, columns = fnames)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# create features for all train data\n",
    "targets = target00001.iloc[:,2:22]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "feats = feats[np.isfinite(targets.sum(axis=1, skipna=False))]\n",
    "targets = targets[np.isfinite(targets.sum(axis=1, skipna=False))]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import xgboost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "ename": "ValueError",
     "evalue": "DataFrame for label cannot have multiple columns",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-54-4c763af0299c>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mxgboost\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mDMatrix\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdata\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mfeats\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlabel\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtargets\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;32m/opt/conda/lib/python3.5/site-packages/xgboost-0.4-py3.5.egg/xgboost/core.py\u001b[0m in \u001b[0;36m__init__\u001b[1;34m(self, data, label, missing, weight, silent, feature_names, feature_types)\u001b[0m\n\u001b[0;32m    220\u001b[0m                                                                 \u001b[0mfeature_names\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    221\u001b[0m                                                                 feature_types)\n\u001b[1;32m--> 222\u001b[1;33m         \u001b[0mlabel\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0m_maybe_pandas_label\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mlabel\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    223\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    224\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdata\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mSTRING_TYPES\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m/opt/conda/lib/python3.5/site-packages/xgboost-0.4-py3.5.egg/xgboost/core.py\u001b[0m in \u001b[0;36m_maybe_pandas_label\u001b[1;34m(label)\u001b[0m\n\u001b[0;32m    164\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mlabel\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mDataFrame\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    165\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mlabel\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcolumns\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m>\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 166\u001b[1;33m             \u001b[1;32mraise\u001b[0m \u001b[0mValueError\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'DataFrame for label cannot have multiple columns'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    167\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    168\u001b[0m         \u001b[0mlabel_dtypes\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mlabel\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdtypes\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mValueError\u001b[0m: DataFrame for label cannot have multiple columns"
     ]
    }
   ],
   "source": [
    "xgboost.DMatrix(data = feats, label = targets)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [Root]",
   "language": "python",
   "name": "Python [Root]"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}