BinaryMatch.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import twosigmafunc\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import seaborn\n",
    "plt.rcParams['figure.figsize'] = (14.0, 8.0) # set default size of plots\n",
    "plt.rcParams['image.interpolation'] = 'nearest'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "train, origin_features = twosigmafunc.preprocess()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "train = train.fillna(train.median())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>derived_0</th>\n",
       "      <th>derived_1</th>\n",
       "      <th>derived_2</th>\n",
       "      <th>derived_3</th>\n",
       "      <th>derived_4</th>\n",
       "      <th>fundamental_0</th>\n",
       "      <th>fundamental_1</th>\n",
       "      <th>fundamental_2</th>\n",
       "      <th>...</th>\n",
       "      <th>technical_36</th>\n",
       "      <th>technical_37</th>\n",
       "      <th>technical_38</th>\n",
       "      <th>technical_39</th>\n",
       "      <th>technical_40</th>\n",
       "      <th>technical_41</th>\n",
       "      <th>technical_42</th>\n",
       "      <th>technical_43</th>\n",
       "      <th>technical_44</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>0.370326</td>\n",
       "      <td>-0.006316</td>\n",
       "      <td>0.222831</td>\n",
       "      <td>-0.213030</td>\n",
       "      <td>0.729277</td>\n",
       "      <td>-0.335633</td>\n",
       "      <td>0.113292</td>\n",
       "      <td>1.621238</td>\n",
       "      <td>...</td>\n",
       "      <td>0.775208</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.414776</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.011753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>11</td>\n",
       "      <td>0</td>\n",
       "      <td>0.014765</td>\n",
       "      <td>-0.038064</td>\n",
       "      <td>-0.017425</td>\n",
       "      <td>0.320652</td>\n",
       "      <td>-0.034134</td>\n",
       "      <td>0.004413</td>\n",
       "      <td>0.114285</td>\n",
       "      <td>-0.210185</td>\n",
       "      <td>...</td>\n",
       "      <td>0.025590</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.273607</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.001240</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.010622</td>\n",
       "      <td>-0.050577</td>\n",
       "      <td>3.379575</td>\n",
       "      <td>-0.157525</td>\n",
       "      <td>-0.068550</td>\n",
       "      <td>-0.155937</td>\n",
       "      <td>1.219439</td>\n",
       "      <td>-0.764516</td>\n",
       "      <td>...</td>\n",
       "      <td>0.151881</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.175710</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.020940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>25</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.178495</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.007262</td>\n",
       "      <td>...</td>\n",
       "      <td>1.035936</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.211506</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.015959</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>26</td>\n",
       "      <td>0</td>\n",
       "      <td>0.176693</td>\n",
       "      <td>-0.025284</td>\n",
       "      <td>-0.057680</td>\n",
       "      <td>0.015100</td>\n",
       "      <td>0.180894</td>\n",
       "      <td>0.139445</td>\n",
       "      <td>-0.125687</td>\n",
       "      <td>-0.018707</td>\n",
       "      <td>...</td>\n",
       "      <td>0.630232</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.001957</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.007338</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 111 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   id  timestamp  derived_0  derived_1  derived_2  derived_3  derived_4  \\\n",
       "0  10          0   0.370326  -0.006316   0.222831  -0.213030   0.729277   \n",
       "1  11          0   0.014765  -0.038064  -0.017425   0.320652  -0.034134   \n",
       "2  12          0  -0.010622  -0.050577   3.379575  -0.157525  -0.068550   \n",
       "3  25          0        NaN        NaN        NaN        NaN        NaN   \n",
       "4  26          0   0.176693  -0.025284  -0.057680   0.015100   0.180894   \n",
       "\n",
       "   fundamental_0  fundamental_1  fundamental_2    ...     technical_36  \\\n",
       "0      -0.335633       0.113292       1.621238    ...         0.775208   \n",
       "1       0.004413       0.114285      -0.210185    ...         0.025590   \n",
       "2      -0.155937       1.219439      -0.764516    ...         0.151881   \n",
       "3       0.178495            NaN      -0.007262    ...         1.035936   \n",
       "4       0.139445      -0.125687      -0.018707    ...         0.630232   \n",
       "\n",
       "   technical_37  technical_38  technical_39  technical_40  technical_41  \\\n",
       "0           NaN           NaN           NaN     -0.414776           NaN   \n",
       "1           NaN           NaN           NaN     -0.273607           NaN   \n",
       "2           NaN           NaN           NaN     -0.175710           NaN   \n",
       "3           NaN           NaN           NaN     -0.211506           NaN   \n",
       "4           NaN           NaN           NaN     -0.001957           NaN   \n",
       "\n",
       "   technical_42  technical_43  technical_44         y  \n",
       "0           NaN          -2.0           NaN -0.011753  \n",
       "1           NaN          -2.0           NaN -0.001240  \n",
       "2           NaN          -2.0           NaN -0.020940  \n",
       "3           NaN          -2.0           NaN -0.015959  \n",
       "4           NaN           0.0           NaN -0.007338  \n",
       "\n",
       "[5 rows x 111 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "gp_t22 = train.groupby('technical_22')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>derived_0</th>\n",
       "      <th>derived_1</th>\n",
       "      <th>derived_2</th>\n",
       "      <th>derived_3</th>\n",
       "      <th>derived_4</th>\n",
       "      <th>fundamental_0</th>\n",
       "      <th>fundamental_1</th>\n",
       "      <th>fundamental_2</th>\n",
       "      <th>...</th>\n",
       "      <th>technical_36</th>\n",
       "      <th>technical_37</th>\n",
       "      <th>technical_38</th>\n",
       "      <th>technical_39</th>\n",
       "      <th>technical_40</th>\n",
       "      <th>technical_41</th>\n",
       "      <th>technical_42</th>\n",
       "      <th>technical_43</th>\n",
       "      <th>technical_44</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>technical_22</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>-0.5</th>\n",
       "      <td>1084.283674</td>\n",
       "      <td>942.896091</td>\n",
       "      <td>0.212794</td>\n",
       "      <td>8.495288e+10</td>\n",
       "      <td>-0.230547</td>\n",
       "      <td>-0.018741</td>\n",
       "      <td>0.016451</td>\n",
       "      <td>-0.078311</td>\n",
       "      <td>-2.221473e+06</td>\n",
       "      <td>0.032147</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.017801</td>\n",
       "      <td>-0.090672</td>\n",
       "      <td>-0.090554</td>\n",
       "      <td>-0.076976</td>\n",
       "      <td>-0.153452</td>\n",
       "      <td>0.004293</td>\n",
       "      <td>-0.026118</td>\n",
       "      <td>-0.825820</td>\n",
       "      <td>-0.000074</td>\n",
       "      <td>0.000252</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.0</th>\n",
       "      <td>1100.609701</td>\n",
       "      <td>947.949630</td>\n",
       "      <td>0.237597</td>\n",
       "      <td>1.282540e+12</td>\n",
       "      <td>0.088750</td>\n",
       "      <td>-0.286017</td>\n",
       "      <td>0.934772</td>\n",
       "      <td>-0.015374</td>\n",
       "      <td>-1.028749e+09</td>\n",
       "      <td>-0.032725</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.036243</td>\n",
       "      <td>-0.094872</td>\n",
       "      <td>-0.081667</td>\n",
       "      <td>-0.074956</td>\n",
       "      <td>-0.002533</td>\n",
       "      <td>0.005093</td>\n",
       "      <td>-0.019153</td>\n",
       "      <td>-0.924458</td>\n",
       "      <td>0.000252</td>\n",
       "      <td>0.000220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5</th>\n",
       "      <td>1096.854701</td>\n",
       "      <td>946.074580</td>\n",
       "      <td>-13.691469</td>\n",
       "      <td>8.519691e+11</td>\n",
       "      <td>-0.613694</td>\n",
       "      <td>-1.092332</td>\n",
       "      <td>40.914593</td>\n",
       "      <td>0.033220</td>\n",
       "      <td>-1.005064e-02</td>\n",
       "      <td>-0.407628</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.206295</td>\n",
       "      <td>-0.084554</td>\n",
       "      <td>-0.069560</td>\n",
       "      <td>-0.064019</td>\n",
       "      <td>0.309784</td>\n",
       "      <td>0.005949</td>\n",
       "      <td>-0.004824</td>\n",
       "      <td>-1.172962</td>\n",
       "      <td>0.000844</td>\n",
       "      <td>0.000193</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 110 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                       id   timestamp  derived_0     derived_1  derived_2  \\\n",
       "technical_22                                                                \n",
       "-0.5          1084.283674  942.896091   0.212794  8.495288e+10  -0.230547   \n",
       " 0.0          1100.609701  947.949630   0.237597  1.282540e+12   0.088750   \n",
       " 0.5          1096.854701  946.074580 -13.691469  8.519691e+11  -0.613694   \n",
       "\n",
       "              derived_3  derived_4  fundamental_0  fundamental_1  \\\n",
       "technical_22                                                       \n",
       "-0.5          -0.018741   0.016451      -0.078311  -2.221473e+06   \n",
       " 0.0          -0.286017   0.934772      -0.015374  -1.028749e+09   \n",
       " 0.5          -1.092332  40.914593       0.033220  -1.005064e-02   \n",
       "\n",
       "              fundamental_2    ...     technical_36  technical_37  \\\n",
       "technical_22                   ...                                  \n",
       "-0.5               0.032147    ...        -0.017801     -0.090672   \n",
       " 0.0              -0.032725    ...        -0.036243     -0.094872   \n",
       " 0.5              -0.407628    ...        -0.206295     -0.084554   \n",
       "\n",
       "              technical_38  technical_39  technical_40  technical_41  \\\n",
       "technical_22                                                           \n",
       "-0.5             -0.090554     -0.076976     -0.153452      0.004293   \n",
       " 0.0             -0.081667     -0.074956     -0.002533      0.005093   \n",
       " 0.5             -0.069560     -0.064019      0.309784      0.005949   \n",
       "\n",
       "              technical_42  technical_43  technical_44         y  \n",
       "technical_22                                                      \n",
       "-0.5             -0.026118     -0.825820     -0.000074  0.000252  \n",
       " 0.0             -0.019153     -0.924458      0.000252  0.000220  \n",
       " 0.5             -0.004824     -1.172962      0.000844  0.000193  \n",
       "\n",
       "[3 rows x 110 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gp_t22.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>derived_0</th>\n",
       "      <th>derived_1</th>\n",
       "      <th>derived_2</th>\n",
       "      <th>derived_3</th>\n",
       "      <th>derived_4</th>\n",
       "      <th>fundamental_0</th>\n",
       "      <th>fundamental_1</th>\n",
       "      <th>fundamental_2</th>\n",
       "      <th>...</th>\n",
       "      <th>technical_36</th>\n",
       "      <th>technical_37</th>\n",
       "      <th>technical_38</th>\n",
       "      <th>technical_39</th>\n",
       "      <th>technical_40</th>\n",
       "      <th>technical_41</th>\n",
       "      <th>technical_42</th>\n",
       "      <th>technical_43</th>\n",
       "      <th>technical_44</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>technical_22</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>-0.5</th>\n",
       "      <td>645.628554</td>\n",
       "      <td>519.521150</td>\n",
       "      <td>10.890799</td>\n",
       "      <td>1.838422e+13</td>\n",
       "      <td>28.879731</td>\n",
       "      <td>18.431871</td>\n",
       "      <td>0.634549</td>\n",
       "      <td>0.219565</td>\n",
       "      <td>4.342770e+08</td>\n",
       "      <td>0.952671</td>\n",
       "      <td>...</td>\n",
       "      <td>0.312316</td>\n",
       "      <td>0.239298</td>\n",
       "      <td>0.240534</td>\n",
       "      <td>0.223088</td>\n",
       "      <td>0.139860</td>\n",
       "      <td>0.111728</td>\n",
       "      <td>0.223019</td>\n",
       "      <td>0.942196</td>\n",
       "      <td>0.028609</td>\n",
       "      <td>0.013988</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.0</th>\n",
       "      <td>624.809010</td>\n",
       "      <td>519.917191</td>\n",
       "      <td>38.554620</td>\n",
       "      <td>1.051663e+14</td>\n",
       "      <td>15.410796</td>\n",
       "      <td>26.497681</td>\n",
       "      <td>76.719482</td>\n",
       "      <td>0.242425</td>\n",
       "      <td>1.008650e+11</td>\n",
       "      <td>2.247143</td>\n",
       "      <td>...</td>\n",
       "      <td>0.486310</td>\n",
       "      <td>0.252411</td>\n",
       "      <td>0.232611</td>\n",
       "      <td>0.225773</td>\n",
       "      <td>0.188392</td>\n",
       "      <td>0.112151</td>\n",
       "      <td>0.204420</td>\n",
       "      <td>0.956095</td>\n",
       "      <td>0.027877</td>\n",
       "      <td>0.019465</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5</th>\n",
       "      <td>621.395693</td>\n",
       "      <td>519.250842</td>\n",
       "      <td>424.450509</td>\n",
       "      <td>7.255277e+13</td>\n",
       "      <td>93.986218</td>\n",
       "      <td>167.042952</td>\n",
       "      <td>1408.642121</td>\n",
       "      <td>0.267388</td>\n",
       "      <td>5.419305e+00</td>\n",
       "      <td>5.110052</td>\n",
       "      <td>...</td>\n",
       "      <td>0.883191</td>\n",
       "      <td>0.247589</td>\n",
       "      <td>0.222450</td>\n",
       "      <td>0.216348</td>\n",
       "      <td>0.352445</td>\n",
       "      <td>0.112023</td>\n",
       "      <td>0.203857</td>\n",
       "      <td>0.952916</td>\n",
       "      <td>0.027391</td>\n",
       "      <td>0.030737</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 110 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                      id   timestamp   derived_0     derived_1  derived_2  \\\n",
       "technical_22                                                                \n",
       "-0.5          645.628554  519.521150   10.890799  1.838422e+13  28.879731   \n",
       " 0.0          624.809010  519.917191   38.554620  1.051663e+14  15.410796   \n",
       " 0.5          621.395693  519.250842  424.450509  7.255277e+13  93.986218   \n",
       "\n",
       "               derived_3    derived_4  fundamental_0  fundamental_1  \\\n",
       "technical_22                                                          \n",
       "-0.5           18.431871     0.634549       0.219565   4.342770e+08   \n",
       " 0.0           26.497681    76.719482       0.242425   1.008650e+11   \n",
       " 0.5          167.042952  1408.642121       0.267388   5.419305e+00   \n",
       "\n",
       "              fundamental_2    ...     technical_36  technical_37  \\\n",
       "technical_22                   ...                                  \n",
       "-0.5               0.952671    ...         0.312316      0.239298   \n",
       " 0.0               2.247143    ...         0.486310      0.252411   \n",
       " 0.5               5.110052    ...         0.883191      0.247589   \n",
       "\n",
       "              technical_38  technical_39  technical_40  technical_41  \\\n",
       "technical_22                                                           \n",
       "-0.5              0.240534      0.223088      0.139860      0.111728   \n",
       " 0.0              0.232611      0.225773      0.188392      0.112151   \n",
       " 0.5              0.222450      0.216348      0.352445      0.112023   \n",
       "\n",
       "              technical_42  technical_43  technical_44         y  \n",
       "technical_22                                                      \n",
       "-0.5              0.223019      0.942196      0.028609  0.013988  \n",
       " 0.0              0.204420      0.956095      0.027877  0.019465  \n",
       " 0.5              0.203857      0.952916      0.027391  0.030737  \n",
       "\n",
       "[3 rows x 110 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gp_t22.std()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Let's see how many category features "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "derived_0\n",
      "derived_1\n",
      "derived_2\n",
      "derived_3\n",
      "derived_4\n",
      "fundamental_0\n",
      "fundamental_1\n",
      "fundamental_2\n",
      "fundamental_3\n",
      "fundamental_5\n",
      "fundamental_6\n",
      "fundamental_7\n",
      "fundamental_8\n",
      "fundamental_9\n",
      "fundamental_10\n",
      "fundamental_11\n",
      "fundamental_12\n",
      "fundamental_13\n",
      "fundamental_14\n",
      "fundamental_15\n",
      "fundamental_16\n",
      "fundamental_17\n",
      "fundamental_18\n",
      "fundamental_19\n",
      "fundamental_20\n",
      "fundamental_21\n",
      "fundamental_22\n",
      "fundamental_23\n",
      "fundamental_24\n",
      "fundamental_25\n",
      "fundamental_26\n",
      "fundamental_27\n",
      "fundamental_28\n",
      "fundamental_29\n",
      "fundamental_30\n",
      "fundamental_31\n",
      "fundamental_32\n",
      "fundamental_33\n",
      "fundamental_34\n",
      "fundamental_35\n",
      "fundamental_36\n",
      "fundamental_37\n",
      "fundamental_38\n",
      "fundamental_39\n",
      "fundamental_40\n",
      "fundamental_41\n",
      "fundamental_42\n",
      "fundamental_43\n",
      "fundamental_44\n",
      "fundamental_45\n",
      "fundamental_46\n",
      "fundamental_47\n",
      "fundamental_48\n",
      "fundamental_49\n",
      "fundamental_50\n",
      "fundamental_51\n",
      "fundamental_52\n",
      "fundamental_53\n",
      "fundamental_54\n",
      "fundamental_55\n",
      "fundamental_56\n",
      "fundamental_57\n",
      "fundamental_58\n",
      "fundamental_59\n",
      "fundamental_60\n",
      "fundamental_61\n",
      "fundamental_62\n",
      "fundamental_63\n",
      "technical_0\n",
      "technical_1\n",
      "technical_2\n",
      "technical_3\n",
      "technical_5\n",
      "technical_6\n",
      "technical_7\n",
      "technical_9\n",
      "technical_10\n",
      "technical_11\n",
      "technical_12\n",
      "technical_13\n",
      "technical_14\n",
      "technical_16\n",
      "technical_17\n",
      "technical_18\n",
      "technical_19\n",
      "technical_20\n",
      "technical_21\n",
      "technical_22\n",
      "technical_24\n",
      "technical_25\n",
      "technical_27\n",
      "technical_28\n",
      "technical_29\n",
      "technical_30\n",
      "technical_31\n",
      "technical_32\n",
      "technical_33\n",
      "technical_34\n",
      "technical_35\n",
      "technical_36\n",
      "technical_37\n",
      "technical_38\n",
      "technical_39\n",
      "technical_40\n",
      "technical_41\n",
      "technical_42\n",
      "technical_43\n",
      "technical_44\n"
     ]
    }
   ],
   "source": [
    "counts = {}\n",
    "for col in origin_features:\n",
    "    counts[col] = len(train[col].unique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "sorted_counts = twosigmafunc.sort_dict(counts, key=lambda x : x[1], reverse=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('technical_34', 3),\n",
       " ('technical_22', 3),\n",
       " ('technical_10', 3323),\n",
       " ('technical_29', 5641),\n",
       " ('technical_14', 10260),\n",
       " ('technical_43', 26461),\n",
       " ('technical_16', 44113),\n",
       " ('technical_9', 44642),\n",
       " ('technical_18', 71103),\n",
       " ('technical_6', 75358)]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sorted_counts[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "seems only `technical_34` and `technical_22` are category features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "gp_cates = train.groupby(['technical_22', 'technical_34'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>derived_0</th>\n",
       "      <th>derived_1</th>\n",
       "      <th>derived_2</th>\n",
       "      <th>derived_3</th>\n",
       "      <th>derived_4</th>\n",
       "      <th>fundamental_0</th>\n",
       "      <th>fundamental_1</th>\n",
       "      <th>fundamental_2</th>\n",
       "      <th>...</th>\n",
       "      <th>technical_36</th>\n",
       "      <th>technical_37</th>\n",
       "      <th>technical_38</th>\n",
       "      <th>technical_39</th>\n",
       "      <th>technical_40</th>\n",
       "      <th>technical_41</th>\n",
       "      <th>technical_42</th>\n",
       "      <th>technical_43</th>\n",
       "      <th>technical_44</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>technical_22</th>\n",
       "      <th>technical_34</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">-0.5</th>\n",
       "      <th>-0.5</th>\n",
       "      <td>1127.355425</td>\n",
       "      <td>860.573179</td>\n",
       "      <td>0.545727</td>\n",
       "      <td>2.290590e+11</td>\n",
       "      <td>-0.009731</td>\n",
       "      <td>0.045468</td>\n",
       "      <td>-0.020500</td>\n",
       "      <td>-0.211127</td>\n",
       "      <td>-6.703287e+06</td>\n",
       "      <td>0.078036</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.016320</td>\n",
       "      <td>-0.101642</td>\n",
       "      <td>-0.101068</td>\n",
       "      <td>-0.086080</td>\n",
       "      <td>-0.189163</td>\n",
       "      <td>0.004297</td>\n",
       "      <td>-0.044077</td>\n",
       "      <td>-0.839089</td>\n",
       "      <td>-0.000811</td>\n",
       "      <td>0.000259</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.0</th>\n",
       "      <td>1083.027118</td>\n",
       "      <td>968.653750</td>\n",
       "      <td>0.014406</td>\n",
       "      <td>2.746714e-01</td>\n",
       "      <td>-1.252684</td>\n",
       "      <td>0.049385</td>\n",
       "      <td>0.023585</td>\n",
       "      <td>-0.076532</td>\n",
       "      <td>-2.533412e+05</td>\n",
       "      <td>0.085722</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.006407</td>\n",
       "      <td>-0.098429</td>\n",
       "      <td>-0.091693</td>\n",
       "      <td>-0.082463</td>\n",
       "      <td>-0.145234</td>\n",
       "      <td>0.005470</td>\n",
       "      <td>-0.022914</td>\n",
       "      <td>-0.818600</td>\n",
       "      <td>0.000760</td>\n",
       "      <td>0.000255</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5</th>\n",
       "      <td>1033.128649</td>\n",
       "      <td>1018.287231</td>\n",
       "      <td>0.015719</td>\n",
       "      <td>6.109132e-02</td>\n",
       "      <td>0.135808</td>\n",
       "      <td>-0.161633</td>\n",
       "      <td>0.047893</td>\n",
       "      <td>0.074045</td>\n",
       "      <td>1.956524e-01</td>\n",
       "      <td>0.029041</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.030254</td>\n",
       "      <td>-0.072998</td>\n",
       "      <td>-0.079753</td>\n",
       "      <td>-0.063284</td>\n",
       "      <td>-0.118341</td>\n",
       "      <td>0.003448</td>\n",
       "      <td>-0.008374</td>\n",
       "      <td>-0.817128</td>\n",
       "      <td>-0.000097</td>\n",
       "      <td>0.000241</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.0</th>\n",
       "      <th>-0.5</th>\n",
       "      <td>1063.278103</td>\n",
       "      <td>968.133330</td>\n",
       "      <td>0.202264</td>\n",
       "      <td>4.448216e+12</td>\n",
       "      <td>0.044041</td>\n",
       "      <td>-0.416674</td>\n",
       "      <td>1.942694</td>\n",
       "      <td>-0.139056</td>\n",
       "      <td>-5.746534e+09</td>\n",
       "      <td>-0.088694</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.106677</td>\n",
       "      <td>-0.116783</td>\n",
       "      <td>-0.086224</td>\n",
       "      <td>-0.090996</td>\n",
       "      <td>-0.025983</td>\n",
       "      <td>0.006536</td>\n",
       "      <td>-0.039513</td>\n",
       "      <td>-1.046430</td>\n",
       "      <td>-0.000383</td>\n",
       "      <td>0.000221</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.0</th>\n",
       "      <td>1151.045387</td>\n",
       "      <td>941.829200</td>\n",
       "      <td>0.547492</td>\n",
       "      <td>1.090023e+00</td>\n",
       "      <td>0.152190</td>\n",
       "      <td>-0.261556</td>\n",
       "      <td>1.565129</td>\n",
       "      <td>-0.035375</td>\n",
       "      <td>4.187698e-02</td>\n",
       "      <td>-0.018904</td>\n",
       "      <td>...</td>\n",
       "      <td>0.008197</td>\n",
       "      <td>-0.091531</td>\n",
       "      <td>-0.087089</td>\n",
       "      <td>-0.076603</td>\n",
       "      <td>-0.013048</td>\n",
       "      <td>0.005087</td>\n",
       "      <td>-0.014491</td>\n",
       "      <td>-0.864735</td>\n",
       "      <td>0.000343</td>\n",
       "      <td>0.000226</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 109 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                    id    timestamp  derived_0     derived_1  \\\n",
       "technical_22 technical_34                                                      \n",
       "-0.5         -0.5          1127.355425   860.573179   0.545727  2.290590e+11   \n",
       "              0.0          1083.027118   968.653750   0.014406  2.746714e-01   \n",
       "              0.5          1033.128649  1018.287231   0.015719  6.109132e-02   \n",
       " 0.0         -0.5          1063.278103   968.133330   0.202264  4.448216e+12   \n",
       "              0.0          1151.045387   941.829200   0.547492  1.090023e+00   \n",
       "\n",
       "                           derived_2  derived_3  derived_4  fundamental_0  \\\n",
       "technical_22 technical_34                                                   \n",
       "-0.5         -0.5          -0.009731   0.045468  -0.020500      -0.211127   \n",
       "              0.0          -1.252684   0.049385   0.023585      -0.076532   \n",
       "              0.5           0.135808  -0.161633   0.047893       0.074045   \n",
       " 0.0         -0.5           0.044041  -0.416674   1.942694      -0.139056   \n",
       "              0.0           0.152190  -0.261556   1.565129      -0.035375   \n",
       "\n",
       "                           fundamental_1  fundamental_2    ...     \\\n",
       "technical_22 technical_34                                  ...      \n",
       "-0.5         -0.5          -6.703287e+06       0.078036    ...      \n",
       "              0.0          -2.533412e+05       0.085722    ...      \n",
       "              0.5           1.956524e-01       0.029041    ...      \n",
       " 0.0         -0.5          -5.746534e+09      -0.088694    ...      \n",
       "              0.0           4.187698e-02      -0.018904    ...      \n",
       "\n",
       "                           technical_36  technical_37  technical_38  \\\n",
       "technical_22 technical_34                                             \n",
       "-0.5         -0.5             -0.016320     -0.101642     -0.101068   \n",
       "              0.0             -0.006407     -0.098429     -0.091693   \n",
       "              0.5             -0.030254     -0.072998     -0.079753   \n",
       " 0.0         -0.5             -0.106677     -0.116783     -0.086224   \n",
       "              0.0              0.008197     -0.091531     -0.087089   \n",
       "\n",
       "                           technical_39  technical_40  technical_41  \\\n",
       "technical_22 technical_34                                             \n",
       "-0.5         -0.5             -0.086080     -0.189163      0.004297   \n",
       "              0.0             -0.082463     -0.145234      0.005470   \n",
       "              0.5             -0.063284     -0.118341      0.003448   \n",
       " 0.0         -0.5             -0.090996     -0.025983      0.006536   \n",
       "              0.0             -0.076603     -0.013048      0.005087   \n",
       "\n",
       "                           technical_42  technical_43  technical_44         y  \n",
       "technical_22 technical_34                                                      \n",
       "-0.5         -0.5             -0.044077     -0.839089     -0.000811  0.000259  \n",
       "              0.0             -0.022914     -0.818600      0.000760  0.000255  \n",
       "              0.5             -0.008374     -0.817128     -0.000097  0.000241  \n",
       " 0.0         -0.5             -0.039513     -1.046430     -0.000383  0.000221  \n",
       "              0.0             -0.014491     -0.864735      0.000343  0.000226  \n",
       "\n",
       "[5 rows x 109 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gp_cates.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Linear Corref in different groups "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.4/site-packages/numpy/lib/function_base.py:3003: RuntimeWarning: invalid value encountered in true_divide\n",
      "  c /= stddev[:, None]\n",
      "/usr/local/lib/python3.4/site-packages/numpy/lib/function_base.py:3004: RuntimeWarning: invalid value encountered in true_divide\n",
      "  c /= stddev[None, :]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing -0.5 -0.5\n",
      "processing -0.5 0\n",
      "processing -0.5 0.5\n",
      "processing 0 -0.5\n",
      "processing 0 0\n",
      "processing 0 0.5\n",
      "processing 0.5 -0.5\n",
      "processing 0.5 0\n",
      "processing 0.5 0.5\n"
     ]
    }
   ],
   "source": [
    "values = [-0.5, 0, 0.5]\n",
    "corref = {}\n",
    "for t22_value in values:\n",
    "    for t34_value in values:\n",
    "        x = train.loc[(train.technical_22 == t22_value) & (train.technical_34 == t34_value)]\n",
    "        corref[(t22_value, t34_value)] = twosigmafunc.feature_corref(x, 'y', origin_features)\n",
    "        print('processing', t22_value, t34_value)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(0.5, -0.5)\n",
      "[('fundamental_14', -0.0088358897004402158), ('fundamental_23', -0.0087673106250023194)]\n",
      "(0.5, 0)\n",
      "[('fundamental_51', 0.011901245177861104), ('derived_4', 0.0074380590633454239)]\n",
      "(0, -0.5)\n",
      "[('technical_20', -0.020661451132074474), ('technical_14', -0.0069656369913574587)]\n",
      "(0, 0.5)\n",
      "[('fundamental_2', 0.0076004432572126272), ('technical_30', 0.031828576173331465)]\n",
      "(0, 0)\n",
      "[('technical_20', -0.018375491495355476), ('technical_30', 0.012091473483376011)]\n",
      "(-0.5, 0)\n",
      "[('technical_20', -0.023838571297583232), ('technical_14', -0.0069090630788274012)]\n",
      "(-0.5, -0.5)\n",
      "[('technical_20', -0.026938698396285009), ('technical_30', 0.016293959725609947)]\n",
      "(0.5, 0.5)\n",
      "[('technical_30', 0.024695513679455283), ('technical_20', -0.022300057631401301)]\n",
      "(-0.5, 0.5)\n",
      "[('fundamental_2', 0.0097173480804507543), ('technical_30', 0.031298077922583517)]\n"
     ]
    }
   ],
   "source": [
    "features_groups = {}\n",
    "for key in corref.keys():\n",
    "    print(key)\n",
    "    best_features = [col for col in corref[key][:5] if col[0] not in ['technical_22', 'technical_34']][:2]\n",
    "    features_groups[key] = [best_features[0][0], best_features[1][0]]\n",
    "    print(best_features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{(-0.5, -0.5): ['technical_20', 'technical_30'],\n",
       " (-0.5, 0): ['technical_20', 'technical_14'],\n",
       " (-0.5, 0.5): ['fundamental_2', 'technical_30'],\n",
       " (0, -0.5): ['technical_20', 'technical_14'],\n",
       " (0, 0): ['technical_20', 'technical_30'],\n",
       " (0, 0.5): ['fundamental_2', 'technical_30'],\n",
       " (0.5, -0.5): ['fundamental_14', 'fundamental_23'],\n",
       " (0.5, 0): ['fundamental_51', 'derived_4'],\n",
       " (0.5, 0.5): ['technical_30', 'technical_20']}"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features_groups"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "lr_features = []\n",
    "for col in features_groups.values():\n",
    "    lr_features.extend(col)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "lr_features = list(set(lr_features))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['derived_4',\n",
       " 'technical_14',\n",
       " 'fundamental_2',\n",
       " 'technical_30',\n",
       " 'fundamental_14',\n",
       " 'fundamental_23',\n",
       " 'fundamental_51',\n",
       " 'technical_20']"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lr_features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Train those linear_models "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "t22_dummy = pd.get_dummies(train.technical_22, prefix='t22')\n",
    "train = pd.concat([train, t22_dummy], axis=1)\n",
    "t34_dummy = pd.get_dummies(train.technical_34, prefix='t34')\n",
    "train = pd.concat([train, t34_dummy], axis=1)\n",
    "train.drop(['t22_0.5', 't34_0.5'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_train, y_train, X_test, y_test = twosigmafunc.split_data(train,  ['t22_0.0', 't34_0.0', 't22_-0.5', 't34_-0.5', \n",
    "                                                                    'timestamp', 'technical_22', 'technical_34'] + lr_features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "low_y_cut = -0.075\n",
    "high_y_cut = 0.075\n",
    "y_is_above_cut = (y_train > high_y_cut)\n",
    "y_is_below_cut = (y_train < low_y_cut)\n",
    "y_is_within_cut = (~y_is_above_cut & ~y_is_below_cut)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>technical_20</th>\n",
       "      <th>technical_12</th>\n",
       "      <th>technical_30</th>\n",
       "      <th>technical_27</th>\n",
       "      <th>technical_35</th>\n",
       "      <th>fundamental_53</th>\n",
       "      <th>fundamental_14</th>\n",
       "      <th>fundamental_11</th>\n",
       "      <th>t22_0.0</th>\n",
       "      <th>t34_0.0</th>\n",
       "      <th>t22_-0.5</th>\n",
       "      <th>t34_-0.5</th>\n",
       "      <th>fundamental_18</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>technical_22</th>\n",
       "      <th>technical_34</th>\n",
       "      <th>(-0.5, -0.5)</th>\n",
       "      <th>(0.5, 0.5)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.000058</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.427469</td>\n",
       "      <td>0.937880</td>\n",
       "      <td>0.208402</td>\n",
       "      <td>-0.055747</td>\n",
       "      <td>1.412742</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.404180</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000373</td>\n",
       "      <td>0.000078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.000058</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.154613</td>\n",
       "      <td>0.232154</td>\n",
       "      <td>-0.281368</td>\n",
       "      <td>-0.034664</td>\n",
       "      <td>-0.202247</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.208746</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.5</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.000373</td>\n",
       "      <td>0.000078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.006942</td>\n",
       "      <td>-0.000058</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.378332</td>\n",
       "      <td>0.372688</td>\n",
       "      <td>-0.026392</td>\n",
       "      <td>-0.031116</td>\n",
       "      <td>-0.899515</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.077977</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.5</td>\n",
       "      <td>-0.5</td>\n",
       "      <td>-0.000835</td>\n",
       "      <td>-0.000537</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.006766</td>\n",
       "      <td>-0.000058</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.023713</td>\n",
       "      <td>0.751021</td>\n",
       "      <td>0.085835</td>\n",
       "      <td>0.018161</td>\n",
       "      <td>0.019866</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.310245</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.5</td>\n",
       "      <td>-0.000805</td>\n",
       "      <td>-0.000521</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.006236</td>\n",
       "      <td>-0.000058</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.741545</td>\n",
       "      <td>0.595206</td>\n",
       "      <td>0.082511</td>\n",
       "      <td>0.139590</td>\n",
       "      <td>-0.167811</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.021833</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.5</td>\n",
       "      <td>0.5</td>\n",
       "      <td>-0.000712</td>\n",
       "      <td>-0.000474</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   technical_20  technical_12  technical_30  technical_27  technical_35  \\\n",
       "0      0.000000     -0.000058           0.0      1.427469      0.937880   \n",
       "1      0.000000     -0.000058           0.0      0.154613      0.232154   \n",
       "2      0.006942     -0.000058           0.0      0.378332      0.372688   \n",
       "3      0.006766     -0.000058           0.0      1.023713      0.751021   \n",
       "4      0.006236     -0.000058           0.0      0.741545      0.595206   \n",
       "\n",
       "   fundamental_53  fundamental_14  fundamental_11  t22_0.0  t34_0.0  t22_-0.5  \\\n",
       "0        0.208402       -0.055747        1.412742        1        1         0   \n",
       "1       -0.281368       -0.034664       -0.202247        0        0         1   \n",
       "2       -0.026392       -0.031116       -0.899515        0        0         1   \n",
       "3        0.085835        0.018161        0.019866        1        0         0   \n",
       "4        0.082511        0.139590       -0.167811        0        0         1   \n",
       "\n",
       "   t34_-0.5  fundamental_18  timestamp  technical_22  technical_34  \\\n",
       "0         0        0.404180          0           0.0           0.0   \n",
       "1         0       -0.208746          0          -0.5           0.5   \n",
       "2         1        0.077977          0          -0.5          -0.5   \n",
       "3         1       -0.310245          0           0.0          -0.5   \n",
       "4         0        0.021833          0          -0.5           0.5   \n",
       "\n",
       "   (-0.5, -0.5)  (0.5, 0.5)  \n",
       "0      0.000373    0.000078  \n",
       "1      0.000373    0.000078  \n",
       "2     -0.000835   -0.000537  \n",
       "3     -0.000805   -0.000521  \n",
       "4     -0.000712   -0.000474  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "features = twosigmafunc.origin_features(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(-0.5, -0.5) ['technical_20', 'technical_30']\n",
      "size: 118789\n",
      "(0.5, 0.5) ['technical_30', 'technical_20']\n",
      "size: 98880\n",
      "(0, 0.5) ['fundamental_2', 'technical_30']\n",
      "size: 93331\n",
      "(-0.5, 0) ['technical_20', 'technical_14']\n",
      "size: 79246\n",
      "(0, 0) ['technical_20', 'technical_30']\n",
      "size: 96501\n",
      "(0.5, 0) ['fundamental_51', 'derived_4']\n",
      "size: 89986\n",
      "(0.5, -0.5) ['fundamental_14', 'fundamental_23']\n",
      "size: 75987\n",
      "(-0.5, 0.5) ['fundamental_2', 'technical_30']\n",
      "size: 75284\n",
      "(0, -0.5) ['technical_20', 'technical_14']\n",
      "size: 78294\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import LinearRegression, Ridge\n",
    "linear_models = []\n",
    "meta_features = []\n",
    "for key, value in features_groups.items():\n",
    "    print(key, value)\n",
    "    t22, t34 = key\n",
    "    ind = X_train.loc[(X_train.technical_22 == t22) & (X_train.technical_34 == t34)].index\n",
    "    \n",
    "    print('size:', len(ind))\n",
    "    meta_name = str(key)\n",
    "    clf = LinearRegression(n_jobs=-1, normalize=True)\n",
    "    clf.fit(X_train.loc[ind, value], y_train.loc[ind])\n",
    "    linear_models.append(clf)\n",
    "    X_train[meta_name] = clf.predict(X_train[value]).clip(low_y_cut, high_y_cut)\n",
    "    X_test[meta_name] = clf.predict(X_test[value]).clip(low_y_cut, high_y_cut)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Let's train stacking model  on linear space"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import ExtraTreesRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "features = [col for col in features if col not in ['technical_22', 'technical_34']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['t22_0.0',\n",
       " 't34_0.0',\n",
       " 't22_-0.5',\n",
       " 't34_-0.5',\n",
       " 'derived_4',\n",
       " 'technical_14',\n",
       " 'fundamental_2',\n",
       " 'technical_30',\n",
       " 'fundamental_14',\n",
       " 'fundamental_23',\n",
       " 'fundamental_51',\n",
       " 'technical_20',\n",
       " '(-0.5, -0.5)',\n",
       " '(0.5, 0.5)',\n",
       " '(0, 0.5)',\n",
       " '(-0.5, 0)',\n",
       " '(0, 0)',\n",
       " '(0.5, 0)',\n",
       " '(0.5, -0.5)',\n",
       " '(-0.5, 0.5)',\n",
       " '(0, -0.5)']"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "etr = ExtraTreesRegressor(n_estimators=100, n_jobs=-1, max_depth=6, max_features=0.9, min_samples_leaf=10, random_state=11)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ExtraTreesRegressor(bootstrap=False, criterion='mse', max_depth=6,\n",
       "          max_features=0.9, max_leaf_nodes=None, min_impurity_split=1e-07,\n",
       "          min_samples_leaf=10, min_samples_split=2,\n",
       "          min_weight_fraction_leaf=0.0, n_estimators=100, n_jobs=-1,\n",
       "          oob_score=False, random_state=11, verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "etr.fit(X_train[features], y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "pred = etr.predict(X_test[features])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.023790084927076998"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "twosigmafunc.R_score(pred, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "imp = pd.DataFrame({'feature': features, 'weight':etr.feature_importances_}).sort_values(\n",
    "    'weight', ascending=False) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>feature</th>\n",
       "      <th>weight</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>(-0.5, -0.5)</td>\n",
       "      <td>0.099785</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>(0.5, 0.5)</td>\n",
       "      <td>0.092121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>(0, 0)</td>\n",
       "      <td>0.079746</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>(-0.5, 0)</td>\n",
       "      <td>0.078091</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>(0, -0.5)</td>\n",
       "      <td>0.077169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>fundamental_23</td>\n",
       "      <td>0.058972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>(0, 0.5)</td>\n",
       "      <td>0.056580</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>t22_0.0</td>\n",
       "      <td>0.056188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>technical_30</td>\n",
       "      <td>0.052951</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>technical_20</td>\n",
       "      <td>0.045098</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           feature    weight\n",
       "12    (-0.5, -0.5)  0.099785\n",
       "13      (0.5, 0.5)  0.092121\n",
       "16          (0, 0)  0.079746\n",
       "15       (-0.5, 0)  0.078091\n",
       "20       (0, -0.5)  0.077169\n",
       "9   fundamental_23  0.058972\n",
       "14        (0, 0.5)  0.056580\n",
       "0          t22_0.0  0.056188\n",
       "7     technical_30  0.052951\n",
       "11    technical_20  0.045098"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "imp[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### See patterns "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "min_stamp = X_test.timestamp.min()\n",
    "max_stamp = X_test.timestamp.max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000 -0.016091030544\n",
      "1100 -0.0168880682811\n",
      "1200 -0.00390829702023\n",
      "1300 -0.00601284643834\n",
      "1400 -0.0211615228029\n",
      "1500 -0.0121171179547\n",
      "1600 0.0292158650679\n",
      "1700 0.02615909934\n",
      "1800 0.0239310928201\n"
     ]
    }
   ],
   "source": [
    "reward = []\n",
    "y_true_acc = []\n",
    "y_pred_acc = []\n",
    "stamp = min_stamp\n",
    "while stamp <= max_stamp:\n",
    "    test = X_test.loc[X_test.timestamp == stamp, features]\n",
    "    y_true = y_test.loc[X_test.timestamp == stamp]\n",
    "    y_pred = etr.predict(test)\n",
    "    y_true_acc.extend(list(y_true))\n",
    "    y_pred_acc.extend(list(y_pred))\n",
    "    reward.append(twosigmafunc.R_score(np.array(y_pred_acc), np.array(y_true_acc)))\n",
    "    stamp += 1\n",
    "    if stamp % 100 == 0:\n",
    "        print(stamp, reward[-1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "score = pd.read_csv('Linear_R_score.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-0.15, 0.07)"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA08AAAHtCAYAAADBdUXWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEEho2KUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x12015c0b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.title('Linear vs Stacking')\n",
    "plt.xlabel('timestamp')\n",
    "plt.ylabel('R score')\n",
    "plt.plot([min_stamp, max_stamp], [0, 0], 'k--')\n",
    "for col in score.columns[1:2]:\n",
    "    plt.plot(range(min_stamp, max_stamp + 1), np.array(score[col]), label=col)\n",
    "plt.plot(range(min_stamp, max_stamp + 1), reward, marker='s', markevery=60, label='Stacking')\n",
    "plt.legend(loc='best')\n",
    "plt.ylim(-0.15, 0.07)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Using Ridge "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(0, 0) ['technical_20', 'technical_35']\n",
      "(0.5, 0) ['fundamental_11', 'technical_20']\n",
      "(0.5, -0.5) ['fundamental_53', 'fundamental_14']\n",
      "(-0.5, 0) ['technical_20', 'technical_30']\n",
      "(-0.5, -0.5) ['technical_20', 'technical_12']\n",
      "(0, -0.5) ['technical_20', 'technical_27']\n"
     ]
    }
   ],
   "source": [
    "linear_models = []\n",
    "meta_features = []\n",
    "for key, value in combinations.items():\n",
    "    print(key, value)\n",
    "    meta_name = str(key)\n",
    "    clf = Ridge(normalize=True)\n",
    "    clf.fit(X_train[value], y_train)\n",
    "    linear_models.append(clf)\n",
    "    X_train[meta_name] = clf.predict(X_train[value]).clip(low_y_cut, high_y_cut)\n",
    "    X_test[meta_name] = clf.predict(X_test[value]).clip(low_y_cut, high_y_cut)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ExtraTreesRegressor(bootstrap=False, criterion='mse', max_depth=6,\n",
       "          max_features=0.9, max_leaf_nodes=None, min_impurity_split=1e-07,\n",
       "          min_samples_leaf=10, min_samples_split=2,\n",
       "          min_weight_fraction_leaf=0.0, n_estimators=100, n_jobs=-1,\n",
       "          oob_score=False, random_state=None, verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "etr.fit(X_train[features], y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.02859741993145842"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred = etr.predict(X_test[features])\n",
    "twosigmafunc.R_score(pred, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000 0.017029952302\n",
      "1100 -0.0131836983879\n",
      "1200 0.00886016414197\n",
      "1300 0.00430314264604\n",
      "1400 -0.0209400293417\n",
      "1500 -0.011698833925\n",
      "1600 0.0311070464829\n",
      "1700 0.030636020112\n",
      "1800 0.0287583661856\n"
     ]
    }
   ],
   "source": [
    "reward_ri = []\n",
    "y_true_acc = []\n",
    "y_pred_acc = []\n",
    "stamp = min_stamp\n",
    "while stamp <= max_stamp:\n",
    "    test = X_test.loc[X_test.timestamp == stamp, features]\n",
    "    y_true = y_test.loc[X_test.timestamp == stamp]\n",
    "    y_pred = etr.predict(test)\n",
    "    y_true_acc.extend(list(y_true))\n",
    "    y_pred_acc.extend(list(y_pred))\n",
    "    reward_ri.append(twosigmafunc.R_score(np.array(y_pred_acc), np.array(y_true_acc)))\n",
    "    stamp += 1\n",
    "    if stamp % 100 == 0:\n",
    "        print(stamp, reward_ri[-1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-0.15, 0.07)"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA08AAAHtCAYAAADBdUXWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nwf8BQThHVYIJm+wAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x12015cb70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.title('Linear vs Stacking')\n",
    "plt.xlabel('timestamp')\n",
    "plt.ylabel('R score')\n",
    "plt.plot([min_stamp, max_stamp], [0, 0], 'k--')\n",
    "for col in score.columns[1:2]:\n",
    "    plt.plot(range(min_stamp, max_stamp + 1), np.array(score[col]), label=col)\n",
    "plt.plot(range(min_stamp, max_stamp + 1), reward, label='Stacking')\n",
    "plt.plot(range(min_stamp, max_stamp + 1), reward_ri, label='Stacking_ridge')\n",
    "plt.legend(loc='best')\n",
    "plt.ylim(-0.15, 0.07)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "no much difference between `linear` and `ridge`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Use 1 feature for linear models "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "combinations = {\n",
    " (-0.5, -0.5): ['technical_20'],\n",
    " (0.5, -0.5): ['fundamental_53'],\n",
    " (0.5, 0): ['fundamental_11'],\n",
    "(-0.5, 0.5): ['technical_30']}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(-0.5, 0.5) ['technical_30']\n",
      "(0.5, 0) ['fundamental_11']\n",
      "(-0.5, -0.5) ['technical_20']\n",
      "(0.5, -0.5) ['fundamental_53']\n"
     ]
    }
   ],
   "source": [
    "linear_models = []\n",
    "meta_features = []\n",
    "for key, value in combinations.items():\n",
    "    print(key, value)\n",
    "    meta_name = str(key)\n",
    "    clf = LinearRegression(n_jobs=-1, normalize=True)\n",
    "    clf.fit(X_train[value].values.reshape(-1, 1), y_train)\n",
    "    linear_models.append(clf)\n",
    "    X_train[meta_name] = clf.predict(X_train[value].values.reshape(-1, 1)).clip(low_y_cut, high_y_cut)\n",
    "    X_test[meta_name] = clf.predict(X_test[value].values.reshape(-1, 1)).clip(low_y_cut, high_y_cut)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['technical_20',\n",
       " 'technical_12',\n",
       " 'technical_30',\n",
       " 'technical_27',\n",
       " 'technical_35',\n",
       " 'fundamental_53',\n",
       " 'fundamental_14',\n",
       " 'fundamental_11',\n",
       " 't22_0.0',\n",
       " 't34_0.0',\n",
       " 't22_-0.5',\n",
       " 't34_-0.5',\n",
       " '(0, 0)',\n",
       " '(0.5, 0)',\n",
       " '(0.5, -0.5)',\n",
       " '(-0.5, 0)',\n",
       " '(-0.5, -0.5)',\n",
       " '(0, -0.5)']"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "features_uni = [\n",
    " 'technical_20',\n",
    " 'technical_30',\n",
    " 'fundamental_53',\n",
    " 'fundamental_11',\n",
    " 't22_0.0',\n",
    " 't34_0.0',\n",
    " 't22_-0.5',\n",
    " 't34_-0.5',\n",
    " '(-0.5, 0.5)',\n",
    " '(0.5, 0)',\n",
    " '(-0.5, -0.5)',\n",
    " '(0.5, -0.5)']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ExtraTreesRegressor(bootstrap=False, criterion='mse', max_depth=6,\n",
       "          max_features=0.9, max_leaf_nodes=None, min_impurity_split=1e-07,\n",
       "          min_samples_leaf=10, min_samples_split=2,\n",
       "          min_weight_fraction_leaf=0.0, n_estimators=100, n_jobs=-1,\n",
       "          oob_score=False, random_state=11, verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "etr.fit(X_train[features_uni], y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.023212708494453158"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred = etr.predict(X_test[features_uni])\n",
    "twosigmafunc.R_score(pred, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000 -0.0131016592031\n",
      "1100 -0.0230794409052\n",
      "1200 -0.0138337371508\n",
      "1300 -0.0141877452261\n",
      "1400 -0.0246030464797\n",
      "1500 -0.0174104294227\n",
      "1600 0.0260206034135\n",
      "1700 0.0251770280351\n",
      "1800 0.0232876497062\n"
     ]
    }
   ],
   "source": [
    "reward_uni = []\n",
    "y_true_acc = []\n",
    "y_pred_acc = []\n",
    "stamp = min_stamp\n",
    "while stamp <= max_stamp:\n",
    "    test = X_test.loc[X_test.timestamp == stamp, features_uni]\n",
    "    y_true = y_test.loc[X_test.timestamp == stamp]\n",
    "    y_pred = etr.predict(test)\n",
    "    y_true_acc.extend(list(y_true))\n",
    "    y_pred_acc.extend(list(y_pred))\n",
    "    reward_uni.append(twosigmafunc.R_score(np.array(y_pred_acc), np.array(y_true_acc)))\n",
    "    stamp += 1\n",
    "    if stamp % 100 == 0:\n",
    "        print(stamp, reward_uni[-1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-0.15, 0.07)"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA08AAAHtCAYAAADBdUXWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11c2d7b70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.title('Linear vs Stacking')\n",
    "plt.xlabel('timestamp')\n",
    "plt.ylabel('R score')\n",
    "plt.plot([min_stamp, max_stamp], [0, 0], 'k--')\n",
    "for col in score.columns[1:2]:\n",
    "    plt.plot(range(min_stamp, max_stamp + 1), np.array(score[col]), label=col)\n",
    "plt.plot(range(min_stamp, max_stamp + 1), reward, label='Stacking')\n",
    "plt.plot(range(min_stamp, max_stamp + 1), reward_ri, label='Stacking_ridge')\n",
    "plt.plot(range(min_stamp, max_stamp + 1), reward_uni, label='Stacking_uni')\n",
    "plt.legend(loc='best')\n",
    "plt.ylim(-0.15, 0.07)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "binary is better!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Add `y_lr_1` and `y_lr_2 `"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "del train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_train['tec20-30'] = X_train.technical_20 - X_train.technical_30\n",
    "X_test['tec20-30'] = X_train.technical_20 - X_train.technical_30"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# See how to fillna with groups "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.technical_22.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.technical_34.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "sub_train = train[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>derived_0</th>\n",
       "      <th>derived_1</th>\n",
       "      <th>derived_2</th>\n",
       "      <th>derived_3</th>\n",
       "      <th>derived_4</th>\n",
       "      <th>fundamental_0</th>\n",
       "      <th>fundamental_1</th>\n",
       "      <th>fundamental_2</th>\n",
       "      <th>...</th>\n",
       "      <th>technical_36</th>\n",
       "      <th>technical_37</th>\n",
       "      <th>technical_38</th>\n",
       "      <th>technical_39</th>\n",
       "      <th>technical_40</th>\n",
       "      <th>technical_41</th>\n",
       "      <th>technical_42</th>\n",
       "      <th>technical_43</th>\n",
       "      <th>technical_44</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>0.370326</td>\n",
       "      <td>-0.006316</td>\n",
       "      <td>0.222831</td>\n",
       "      <td>-0.213030</td>\n",
       "      <td>0.729277</td>\n",
       "      <td>-0.335633</td>\n",
       "      <td>0.113292</td>\n",
       "      <td>1.621238</td>\n",
       "      <td>...</td>\n",
       "      <td>0.775208</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.414776</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.011753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>11</td>\n",
       "      <td>0</td>\n",
       "      <td>0.014765</td>\n",
       "      <td>-0.038064</td>\n",
       "      <td>-0.017425</td>\n",
       "      <td>0.320652</td>\n",
       "      <td>-0.034134</td>\n",
       "      <td>0.004413</td>\n",
       "      <td>0.114285</td>\n",
       "      <td>-0.210185</td>\n",
       "      <td>...</td>\n",
       "      <td>0.025590</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.273607</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.001240</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.010622</td>\n",
       "      <td>-0.050577</td>\n",
       "      <td>3.379575</td>\n",
       "      <td>-0.157525</td>\n",
       "      <td>-0.068550</td>\n",
       "      <td>-0.155937</td>\n",
       "      <td>1.219439</td>\n",
       "      <td>-0.764516</td>\n",
       "      <td>...</td>\n",
       "      <td>0.151881</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.175710</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.020940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>25</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.178495</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.007262</td>\n",
       "      <td>...</td>\n",
       "      <td>1.035936</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.211506</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.015959</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>26</td>\n",
       "      <td>0</td>\n",
       "      <td>0.176693</td>\n",
       "      <td>-0.025284</td>\n",
       "      <td>-0.057680</td>\n",
       "      <td>0.015100</td>\n",
       "      <td>0.180894</td>\n",
       "      <td>0.139445</td>\n",
       "      <td>-0.125687</td>\n",
       "      <td>-0.018707</td>\n",
       "      <td>...</td>\n",
       "      <td>0.630232</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.001957</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.007338</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 111 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   id  timestamp  derived_0  derived_1  derived_2  derived_3  derived_4  \\\n",
       "0  10          0   0.370326  -0.006316   0.222831  -0.213030   0.729277   \n",
       "1  11          0   0.014765  -0.038064  -0.017425   0.320652  -0.034134   \n",
       "2  12          0  -0.010622  -0.050577   3.379575  -0.157525  -0.068550   \n",
       "3  25          0        NaN        NaN        NaN        NaN        NaN   \n",
       "4  26          0   0.176693  -0.025284  -0.057680   0.015100   0.180894   \n",
       "\n",
       "   fundamental_0  fundamental_1  fundamental_2    ...     technical_36  \\\n",
       "0      -0.335633       0.113292       1.621238    ...         0.775208   \n",
       "1       0.004413       0.114285      -0.210185    ...         0.025590   \n",
       "2      -0.155937       1.219439      -0.764516    ...         0.151881   \n",
       "3       0.178495            NaN      -0.007262    ...         1.035936   \n",
       "4       0.139445      -0.125687      -0.018707    ...         0.630232   \n",
       "\n",
       "   technical_37  technical_38  technical_39  technical_40  technical_41  \\\n",
       "0           NaN           NaN           NaN     -0.414776           NaN   \n",
       "1           NaN           NaN           NaN     -0.273607           NaN   \n",
       "2           NaN           NaN           NaN     -0.175710           NaN   \n",
       "3           NaN           NaN           NaN     -0.211506           NaN   \n",
       "4           NaN           NaN           NaN     -0.001957           NaN   \n",
       "\n",
       "   technical_42  technical_43  technical_44         y  \n",
       "0           NaN          -2.0           NaN -0.011753  \n",
       "1           NaN          -2.0           NaN -0.001240  \n",
       "2           NaN          -2.0           NaN -0.020940  \n",
       "3           NaN          -2.0           NaN -0.015959  \n",
       "4           NaN           0.0           NaN -0.007338  \n",
       "\n",
       "[5 rows x 111 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sub_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[  1.00000000e+01,   0.00000000e+00,   3.70326221e-01, ...,\n",
       "         -2.00000000e+00,              nan,  -1.17534492e-02],\n",
       "       [  1.10000000e+01,   0.00000000e+00,   1.47646843e-02, ...,\n",
       "         -2.00000000e+00,              nan,  -1.24013668e-03],\n",
       "       [  1.20000000e+01,   0.00000000e+00,  -1.06217954e-02, ...,\n",
       "         -2.00000000e+00,              nan,  -2.09395438e-02],\n",
       "       ..., \n",
       "       [  3.80000000e+01,   0.00000000e+00,   3.00062418e-01, ...,\n",
       "         -2.00000000e+00,              nan,   1.58029143e-02],\n",
       "       [  3.90000000e+01,   0.00000000e+00,  -3.51085421e-03, ...,\n",
       "         -2.00000000e+00,              nan,  -2.75933668e-02],\n",
       "       [  4.00000000e+01,   0.00000000e+00,  -8.33296925e-02, ...,\n",
       "         -2.00000000e+00,              nan,   6.66242279e-03]], dtype=float32)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sub_train.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "derived_0\n",
      "derived_1\n",
      "derived_2\n",
      "derived_3\n",
      "derived_4\n",
      "fundamental_0\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.4/site-packages/pandas/core/generic.py:3295: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  self._update_inplace(new_data)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "fundamental_1\n",
      "fundamental_2\n",
      "fundamental_3\n",
      "fundamental_5\n",
      "fundamental_6\n",
      "fundamental_7\n",
      "fundamental_8\n",
      "fundamental_9\n",
      "fundamental_10\n",
      "fundamental_11\n",
      "fundamental_12\n",
      "fundamental_13\n",
      "fundamental_14\n",
      "fundamental_15\n",
      "fundamental_16\n",
      "fundamental_17\n",
      "fundamental_18\n",
      "fundamental_19\n",
      "fundamental_20\n",
      "fundamental_21\n",
      "fundamental_22\n",
      "fundamental_23\n",
      "fundamental_24\n",
      "fundamental_25\n",
      "fundamental_26\n",
      "fundamental_27\n",
      "fundamental_28\n",
      "fundamental_29\n",
      "fundamental_30\n",
      "fundamental_31\n",
      "fundamental_32\n",
      "fundamental_33\n",
      "fundamental_34\n",
      "fundamental_35\n",
      "fundamental_36\n",
      "fundamental_37\n",
      "fundamental_38\n",
      "fundamental_39\n",
      "fundamental_40\n",
      "fundamental_41\n",
      "fundamental_42\n",
      "fundamental_43\n",
      "fundamental_44\n",
      "fundamental_45\n",
      "fundamental_46\n",
      "fundamental_47\n",
      "fundamental_48\n",
      "fundamental_49\n",
      "fundamental_50\n",
      "fundamental_51\n",
      "fundamental_52\n",
      "fundamental_53\n",
      "fundamental_54\n",
      "fundamental_55\n",
      "fundamental_56\n",
      "fundamental_57\n",
      "fundamental_58\n",
      "fundamental_59\n",
      "fundamental_60\n",
      "fundamental_61\n",
      "fundamental_62\n",
      "fundamental_63\n",
      "technical_0\n",
      "technical_1\n",
      "technical_2\n",
      "technical_3\n",
      "technical_5\n",
      "technical_6\n",
      "technical_7\n",
      "technical_9\n",
      "technical_10\n",
      "technical_11\n",
      "technical_12\n",
      "technical_13\n",
      "technical_14\n",
      "technical_16\n",
      "technical_17\n",
      "technical_18\n",
      "technical_19\n",
      "technical_20\n",
      "technical_21\n",
      "technical_22\n",
      "technical_24\n",
      "technical_25\n",
      "technical_27\n",
      "technical_28\n",
      "technical_29\n",
      "technical_30\n",
      "technical_31\n",
      "technical_32\n",
      "technical_33\n",
      "technical_34\n",
      "technical_35\n",
      "technical_36\n",
      "technical_37\n",
      "technical_38\n",
      "technical_39\n",
      "technical_40\n",
      "technical_41\n",
      "technical_42\n",
      "technical_43\n",
      "technical_44\n"
     ]
    }
   ],
   "source": [
    "for col in origin_features:\n",
    "    sub_train[col].fillna(sub_train.groupby(['technical_34'])[col].transform(\"mean\"), inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>derived_0</th>\n",
       "      <th>derived_1</th>\n",
       "      <th>derived_2</th>\n",
       "      <th>derived_3</th>\n",
       "      <th>derived_4</th>\n",
       "      <th>fundamental_0</th>\n",
       "      <th>fundamental_1</th>\n",
       "      <th>fundamental_2</th>\n",
       "      <th>...</th>\n",
       "      <th>technical_36</th>\n",
       "      <th>technical_37</th>\n",
       "      <th>technical_38</th>\n",
       "      <th>technical_39</th>\n",
       "      <th>technical_40</th>\n",
       "      <th>technical_41</th>\n",
       "      <th>technical_42</th>\n",
       "      <th>technical_43</th>\n",
       "      <th>technical_44</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>0.370326</td>\n",
       "      <td>-0.006316</td>\n",
       "      <td>0.222831</td>\n",
       "      <td>-0.213030</td>\n",
       "      <td>0.729277</td>\n",
       "      <td>-0.335633</td>\n",
       "      <td>0.113292</td>\n",
       "      <td>1.621238</td>\n",
       "      <td>...</td>\n",
       "      <td>0.775208</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.414776</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.011753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>11</td>\n",
       "      <td>0</td>\n",
       "      <td>0.014765</td>\n",
       "      <td>-0.038064</td>\n",
       "      <td>-0.017425</td>\n",
       "      <td>0.320652</td>\n",
       "      <td>-0.034134</td>\n",
       "      <td>0.004413</td>\n",
       "      <td>0.114285</td>\n",
       "      <td>-0.210185</td>\n",
       "      <td>...</td>\n",
       "      <td>0.025590</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.273607</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.001240</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.010622</td>\n",
       "      <td>-0.050577</td>\n",
       "      <td>3.379575</td>\n",
       "      <td>-0.157525</td>\n",
       "      <td>-0.068550</td>\n",
       "      <td>-0.155937</td>\n",
       "      <td>1.219439</td>\n",
       "      <td>-0.764516</td>\n",
       "      <td>...</td>\n",
       "      <td>0.151881</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.175710</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.020940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>25</td>\n",
       "      <td>0</td>\n",
       "      <td>0.138242</td>\n",
       "      <td>0.067212</td>\n",
       "      <td>-1.206004</td>\n",
       "      <td>-0.355480</td>\n",
       "      <td>-0.261329</td>\n",
       "      <td>0.178495</td>\n",
       "      <td>0.379278</td>\n",
       "      <td>-0.007262</td>\n",
       "      <td>...</td>\n",
       "      <td>1.035936</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.211506</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.015959</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>26</td>\n",
       "      <td>0</td>\n",
       "      <td>0.176693</td>\n",
       "      <td>-0.025284</td>\n",
       "      <td>-0.057680</td>\n",
       "      <td>0.015100</td>\n",
       "      <td>0.180894</td>\n",
       "      <td>0.139445</td>\n",
       "      <td>-0.125687</td>\n",
       "      <td>-0.018707</td>\n",
       "      <td>...</td>\n",
       "      <td>0.630232</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.001957</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.007338</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>27</td>\n",
       "      <td>0</td>\n",
       "      <td>0.346856</td>\n",
       "      <td>0.166239</td>\n",
       "      <td>-6.080701</td>\n",
       "      <td>-0.992249</td>\n",
       "      <td>-0.125916</td>\n",
       "      <td>0.345812</td>\n",
       "      <td>0.379278</td>\n",
       "      <td>-0.584239</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.379366</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.001957</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.031425</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>31</td>\n",
       "      <td>0</td>\n",
       "      <td>0.072036</td>\n",
       "      <td>0.014931</td>\n",
       "      <td>0.152601</td>\n",
       "      <td>0.014063</td>\n",
       "      <td>0.351763</td>\n",
       "      <td>-0.193205</td>\n",
       "      <td>0.188372</td>\n",
       "      <td>0.690013</td>\n",
       "      <td>...</td>\n",
       "      <td>1.232558</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.350685</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.032895</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>38</td>\n",
       "      <td>0</td>\n",
       "      <td>0.300062</td>\n",
       "      <td>0.071251</td>\n",
       "      <td>-0.074451</td>\n",
       "      <td>-0.065292</td>\n",
       "      <td>-0.011286</td>\n",
       "      <td>0.026365</td>\n",
       "      <td>0.210249</td>\n",
       "      <td>0.167494</td>\n",
       "      <td>...</td>\n",
       "      <td>0.242646</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.168077</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.015803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>39</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.003511</td>\n",
       "      <td>-0.034270</td>\n",
       "      <td>0.082372</td>\n",
       "      <td>-0.023937</td>\n",
       "      <td>-0.025750</td>\n",
       "      <td>0.007815</td>\n",
       "      <td>0.263451</td>\n",
       "      <td>-0.241212</td>\n",
       "      <td>...</td>\n",
       "      <td>0.186175</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.300842</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.027593</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>40</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.083330</td>\n",
       "      <td>0.081935</td>\n",
       "      <td>-2.048438</td>\n",
       "      <td>-0.206856</td>\n",
       "      <td>-0.839563</td>\n",
       "      <td>-0.234100</td>\n",
       "      <td>-0.291853</td>\n",
       "      <td>-5.722336</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.021965</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.386579</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.006662</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 111 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   id  timestamp  derived_0  derived_1  derived_2  derived_3  derived_4  \\\n",
       "0  10          0   0.370326  -0.006316   0.222831  -0.213030   0.729277   \n",
       "1  11          0   0.014765  -0.038064  -0.017425   0.320652  -0.034134   \n",
       "2  12          0  -0.010622  -0.050577   3.379575  -0.157525  -0.068550   \n",
       "3  25          0   0.138242   0.067212  -1.206004  -0.355480  -0.261329   \n",
       "4  26          0   0.176693  -0.025284  -0.057680   0.015100   0.180894   \n",
       "5  27          0   0.346856   0.166239  -6.080701  -0.992249  -0.125916   \n",
       "6  31          0   0.072036   0.014931   0.152601   0.014063   0.351763   \n",
       "7  38          0   0.300062   0.071251  -0.074451  -0.065292  -0.011286   \n",
       "8  39          0  -0.003511  -0.034270   0.082372  -0.023937  -0.025750   \n",
       "9  40          0  -0.083330   0.081935  -2.048438  -0.206856  -0.839563   \n",
       "\n",
       "   fundamental_0  fundamental_1  fundamental_2    ...     technical_36  \\\n",
       "0      -0.335633       0.113292       1.621238    ...         0.775208   \n",
       "1       0.004413       0.114285      -0.210185    ...         0.025590   \n",
       "2      -0.155937       1.219439      -0.764516    ...         0.151881   \n",
       "3       0.178495       0.379278      -0.007262    ...         1.035936   \n",
       "4       0.139445      -0.125687      -0.018707    ...         0.630232   \n",
       "5       0.345812       0.379278      -0.584239    ...        -0.379366   \n",
       "6      -0.193205       0.188372       0.690013    ...         1.232558   \n",
       "7       0.026365       0.210249       0.167494    ...         0.242646   \n",
       "8       0.007815       0.263451      -0.241212    ...         0.186175   \n",
       "9      -0.234100      -0.291853      -5.722336    ...        -0.021965   \n",
       "\n",
       "   technical_37  technical_38  technical_39  technical_40  technical_41  \\\n",
       "0           NaN           NaN           NaN     -0.414776           NaN   \n",
       "1           NaN           NaN           NaN     -0.273607           NaN   \n",
       "2           NaN           NaN           NaN     -0.175710           NaN   \n",
       "3           NaN           NaN           NaN     -0.211506           NaN   \n",
       "4           NaN           NaN           NaN     -0.001957           NaN   \n",
       "5           NaN           NaN           NaN     -0.001957           NaN   \n",
       "6           NaN           NaN           NaN      0.350685           NaN   \n",
       "7           NaN           NaN           NaN     -0.168077           NaN   \n",
       "8           NaN           NaN           NaN     -0.300842           NaN   \n",
       "9           NaN           NaN           NaN     -0.386579           NaN   \n",
       "\n",
       "   technical_42  technical_43  technical_44         y  \n",
       "0           NaN          -2.0           NaN -0.011753  \n",
       "1           NaN          -2.0           NaN -0.001240  \n",
       "2           NaN          -2.0           NaN -0.020940  \n",
       "3           NaN          -2.0           NaN -0.015959  \n",
       "4           NaN           0.0           NaN -0.007338  \n",
       "5           NaN          -2.0           NaN  0.031425  \n",
       "6           NaN          -2.0           NaN -0.032895  \n",
       "7           NaN          -2.0           NaN  0.015803  \n",
       "8           NaN          -2.0           NaN -0.027593  \n",
       "9           NaN          -2.0           NaN  0.006662  \n",
       "\n",
       "[10 rows x 111 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sub_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>derived_0</th>\n",
       "      <th>derived_1</th>\n",
       "      <th>derived_2</th>\n",
       "      <th>derived_3</th>\n",
       "      <th>derived_4</th>\n",
       "      <th>fundamental_0</th>\n",
       "      <th>fundamental_1</th>\n",
       "      <th>fundamental_2</th>\n",
       "      <th>...</th>\n",
       "      <th>technical_36</th>\n",
       "      <th>technical_37</th>\n",
       "      <th>technical_38</th>\n",
       "      <th>technical_39</th>\n",
       "      <th>technical_40</th>\n",
       "      <th>technical_41</th>\n",
       "      <th>technical_42</th>\n",
       "      <th>technical_43</th>\n",
       "      <th>technical_44</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>0.370326</td>\n",
       "      <td>-0.006316</td>\n",
       "      <td>0.222831</td>\n",
       "      <td>-0.213030</td>\n",
       "      <td>0.729277</td>\n",
       "      <td>-0.335633</td>\n",
       "      <td>0.113292</td>\n",
       "      <td>1.621238</td>\n",
       "      <td>...</td>\n",
       "      <td>0.775208</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.414776</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.011753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>11</td>\n",
       "      <td>0</td>\n",
       "      <td>0.014765</td>\n",
       "      <td>-0.038064</td>\n",
       "      <td>-0.017425</td>\n",
       "      <td>0.320652</td>\n",
       "      <td>-0.034134</td>\n",
       "      <td>0.004413</td>\n",
       "      <td>0.114285</td>\n",
       "      <td>-0.210185</td>\n",
       "      <td>...</td>\n",
       "      <td>0.025590</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.273607</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.001240</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.010622</td>\n",
       "      <td>-0.050577</td>\n",
       "      <td>3.379575</td>\n",
       "      <td>-0.157525</td>\n",
       "      <td>-0.068550</td>\n",
       "      <td>-0.155937</td>\n",
       "      <td>1.219439</td>\n",
       "      <td>-0.764516</td>\n",
       "      <td>...</td>\n",
       "      <td>0.151881</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.175710</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.020940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>25</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.178495</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.007262</td>\n",
       "      <td>...</td>\n",
       "      <td>1.035936</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.211506</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.015959</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>26</td>\n",
       "      <td>0</td>\n",
       "      <td>0.176693</td>\n",
       "      <td>-0.025284</td>\n",
       "      <td>-0.057680</td>\n",
       "      <td>0.015100</td>\n",
       "      <td>0.180894</td>\n",
       "      <td>0.139445</td>\n",
       "      <td>-0.125687</td>\n",
       "      <td>-0.018707</td>\n",
       "      <td>...</td>\n",
       "      <td>0.630232</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.001957</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.007338</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>27</td>\n",
       "      <td>0</td>\n",
       "      <td>0.346856</td>\n",
       "      <td>0.166239</td>\n",
       "      <td>-6.080701</td>\n",
       "      <td>-0.992249</td>\n",
       "      <td>-0.125916</td>\n",
       "      <td>0.345812</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.584239</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.379366</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.001957</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.031425</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>31</td>\n",
       "      <td>0</td>\n",
       "      <td>0.072036</td>\n",
       "      <td>0.014931</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.014063</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.193205</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>1.232558</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.350685</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.032895</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>38</td>\n",
       "      <td>0</td>\n",
       "      <td>0.300062</td>\n",
       "      <td>0.071251</td>\n",
       "      <td>-0.074451</td>\n",
       "      <td>-0.065292</td>\n",
       "      <td>-0.011286</td>\n",
       "      <td>0.026365</td>\n",
       "      <td>0.210249</td>\n",
       "      <td>0.167494</td>\n",
       "      <td>...</td>\n",
       "      <td>0.242646</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.168077</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.015803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>39</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.003511</td>\n",
       "      <td>-0.034270</td>\n",
       "      <td>0.082372</td>\n",
       "      <td>-0.023937</td>\n",
       "      <td>-0.025750</td>\n",
       "      <td>0.007815</td>\n",
       "      <td>0.263451</td>\n",
       "      <td>-0.241212</td>\n",
       "      <td>...</td>\n",
       "      <td>0.186175</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.300842</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.027593</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>40</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.083330</td>\n",
       "      <td>0.081935</td>\n",
       "      <td>-2.048438</td>\n",
       "      <td>-0.206856</td>\n",
       "      <td>-0.839563</td>\n",
       "      <td>-0.234100</td>\n",
       "      <td>-0.291853</td>\n",
       "      <td>-5.722336</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.021965</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.386579</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.006662</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 111 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   id  timestamp  derived_0  derived_1  derived_2  derived_3  derived_4  \\\n",
       "0  10          0   0.370326  -0.006316   0.222831  -0.213030   0.729277   \n",
       "1  11          0   0.014765  -0.038064  -0.017425   0.320652  -0.034134   \n",
       "2  12          0  -0.010622  -0.050577   3.379575  -0.157525  -0.068550   \n",
       "3  25          0        NaN        NaN        NaN        NaN        NaN   \n",
       "4  26          0   0.176693  -0.025284  -0.057680   0.015100   0.180894   \n",
       "5  27          0   0.346856   0.166239  -6.080701  -0.992249  -0.125916   \n",
       "6  31          0   0.072036   0.014931        NaN   0.014063        NaN   \n",
       "7  38          0   0.300062   0.071251  -0.074451  -0.065292  -0.011286   \n",
       "8  39          0  -0.003511  -0.034270   0.082372  -0.023937  -0.025750   \n",
       "9  40          0  -0.083330   0.081935  -2.048438  -0.206856  -0.839563   \n",
       "\n",
       "   fundamental_0  fundamental_1  fundamental_2    ...     technical_36  \\\n",
       "0      -0.335633       0.113292       1.621238    ...         0.775208   \n",
       "1       0.004413       0.114285      -0.210185    ...         0.025590   \n",
       "2      -0.155937       1.219439      -0.764516    ...         0.151881   \n",
       "3       0.178495            NaN      -0.007262    ...         1.035936   \n",
       "4       0.139445      -0.125687      -0.018707    ...         0.630232   \n",
       "5       0.345812            NaN      -0.584239    ...        -0.379366   \n",
       "6      -0.193205            NaN            NaN    ...         1.232558   \n",
       "7       0.026365       0.210249       0.167494    ...         0.242646   \n",
       "8       0.007815       0.263451      -0.241212    ...         0.186175   \n",
       "9      -0.234100      -0.291853      -5.722336    ...        -0.021965   \n",
       "\n",
       "   technical_37  technical_38  technical_39  technical_40  technical_41  \\\n",
       "0           NaN           NaN           NaN     -0.414776           NaN   \n",
       "1           NaN           NaN           NaN     -0.273607           NaN   \n",
       "2           NaN           NaN           NaN     -0.175710           NaN   \n",
       "3           NaN           NaN           NaN     -0.211506           NaN   \n",
       "4           NaN           NaN           NaN     -0.001957           NaN   \n",
       "5           NaN           NaN           NaN     -0.001957           NaN   \n",
       "6           NaN           NaN           NaN      0.350685           NaN   \n",
       "7           NaN           NaN           NaN     -0.168077           NaN   \n",
       "8           NaN           NaN           NaN     -0.300842           NaN   \n",
       "9           NaN           NaN           NaN     -0.386579           NaN   \n",
       "\n",
       "   technical_42  technical_43  technical_44         y  \n",
       "0           NaN          -2.0           NaN -0.011753  \n",
       "1           NaN          -2.0           NaN -0.001240  \n",
       "2           NaN          -2.0           NaN -0.020940  \n",
       "3           NaN          -2.0           NaN -0.015959  \n",
       "4           NaN           0.0           NaN -0.007338  \n",
       "5           NaN          -2.0           NaN  0.031425  \n",
       "6           NaN          -2.0           NaN -0.032895  \n",
       "7           NaN          -2.0           NaN  0.015803  \n",
       "8           NaN          -2.0           NaN -0.027593  \n",
       "9           NaN          -2.0           NaN  0.006662  \n",
       "\n",
       "[10 rows x 111 columns]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "seems only work with one groupby col"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# How long does it take on whole data? "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import timeit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "t0 = timeit.timeit\n",
    "for col in origin_features:\n",
    "    train[col] = train[col].fillna(gp_t34[col].transform(\"mean\"))\n",
    "t1 = timeit.timeit"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "in a minite"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "gp_t22 = train.groupby('technical_22')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "gp_t34 = train.groupby('technical_34')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "t22_larger = gp_t22.std().mean() - gp_t34.std().mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "t22_feature = t22_larger[t22_larger < 0].index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "t34_feature = t22_larger.drop(t22_feature).index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "t22_feature = [col for col in t22_feature if col not in ['technical_22', 'technical_34', 'id', 'timestamp', 'y']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "t34_feature = [col for col in t34_feature if col not in ['technical_22', 'technical_34', 'id', 'timestamp', 'y']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['derived_2',\n",
       " 'derived_3',\n",
       " 'fundamental_0',\n",
       " 'fundamental_1',\n",
       " 'fundamental_12',\n",
       " 'fundamental_24',\n",
       " 'fundamental_25',\n",
       " 'fundamental_28',\n",
       " 'fundamental_3',\n",
       " 'fundamental_32',\n",
       " 'fundamental_33',\n",
       " 'fundamental_36',\n",
       " 'fundamental_38',\n",
       " 'fundamental_39',\n",
       " 'fundamental_41',\n",
       " 'fundamental_44',\n",
       " 'fundamental_46',\n",
       " 'fundamental_49',\n",
       " 'fundamental_56',\n",
       " 'fundamental_58',\n",
       " 'fundamental_61',\n",
       " 'fundamental_7',\n",
       " 'fundamental_8',\n",
       " 'technical_0',\n",
       " 'technical_12',\n",
       " 'technical_18',\n",
       " 'technical_25',\n",
       " 'technical_28',\n",
       " 'technical_3',\n",
       " 'technical_31',\n",
       " 'technical_32',\n",
       " 'technical_33',\n",
       " 'technical_37',\n",
       " 'technical_39',\n",
       " 'technical_42',\n",
       " 'technical_5']"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t34_feature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['derived_0',\n",
       " 'derived_1',\n",
       " 'derived_4',\n",
       " 'fundamental_10',\n",
       " 'fundamental_11',\n",
       " 'fundamental_13',\n",
       " 'fundamental_14',\n",
       " 'fundamental_15',\n",
       " 'fundamental_16',\n",
       " 'fundamental_17',\n",
       " 'fundamental_18',\n",
       " 'fundamental_19',\n",
       " 'fundamental_2',\n",
       " 'fundamental_20',\n",
       " 'fundamental_21',\n",
       " 'fundamental_22',\n",
       " 'fundamental_23',\n",
       " 'fundamental_26',\n",
       " 'fundamental_27',\n",
       " 'fundamental_29',\n",
       " 'fundamental_30',\n",
       " 'fundamental_31',\n",
       " 'fundamental_34',\n",
       " 'fundamental_35',\n",
       " 'fundamental_37',\n",
       " 'fundamental_40',\n",
       " 'fundamental_42',\n",
       " 'fundamental_43',\n",
       " 'fundamental_45',\n",
       " 'fundamental_47',\n",
       " 'fundamental_48',\n",
       " 'fundamental_5',\n",
       " 'fundamental_50',\n",
       " 'fundamental_51',\n",
       " 'fundamental_52',\n",
       " 'fundamental_53',\n",
       " 'fundamental_54',\n",
       " 'fundamental_55',\n",
       " 'fundamental_57',\n",
       " 'fundamental_59',\n",
       " 'fundamental_6',\n",
       " 'fundamental_60',\n",
       " 'fundamental_62',\n",
       " 'fundamental_63',\n",
       " 'fundamental_9',\n",
       " 'technical_1',\n",
       " 'technical_10',\n",
       " 'technical_11',\n",
       " 'technical_13',\n",
       " 'technical_14',\n",
       " 'technical_16',\n",
       " 'technical_17',\n",
       " 'technical_19',\n",
       " 'technical_2',\n",
       " 'technical_20',\n",
       " 'technical_21',\n",
       " 'technical_24',\n",
       " 'technical_27',\n",
       " 'technical_29',\n",
       " 'technical_30',\n",
       " 'technical_35',\n",
       " 'technical_36',\n",
       " 'technical_38',\n",
       " 'technical_40',\n",
       " 'technical_41',\n",
       " 'technical_43',\n",
       " 'technical_44',\n",
       " 'technical_6',\n",
       " 'technical_7',\n",
       " 'technical_9']"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t22_feature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>derived_0</th>\n",
       "      <th>derived_1</th>\n",
       "      <th>derived_2</th>\n",
       "      <th>derived_3</th>\n",
       "      <th>derived_4</th>\n",
       "      <th>fundamental_0</th>\n",
       "      <th>fundamental_1</th>\n",
       "      <th>fundamental_2</th>\n",
       "      <th>...</th>\n",
       "      <th>technical_36</th>\n",
       "      <th>technical_37</th>\n",
       "      <th>technical_38</th>\n",
       "      <th>technical_39</th>\n",
       "      <th>technical_40</th>\n",
       "      <th>technical_41</th>\n",
       "      <th>technical_42</th>\n",
       "      <th>technical_43</th>\n",
       "      <th>technical_44</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>technical_22</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>-0.5</th>\n",
       "      <td>645.628554</td>\n",
       "      <td>519.521150</td>\n",
       "      <td>10.938387</td>\n",
       "      <td>1.838488e+13</td>\n",
       "      <td>28.881748</td>\n",
       "      <td>18.432194</td>\n",
       "      <td>17.259764</td>\n",
       "      <td>0.219686</td>\n",
       "      <td>6.027424e+08</td>\n",
       "      <td>0.959341</td>\n",
       "      <td>...</td>\n",
       "      <td>0.312348</td>\n",
       "      <td>0.239122</td>\n",
       "      <td>0.240359</td>\n",
       "      <td>0.222952</td>\n",
       "      <td>0.139891</td>\n",
       "      <td>0.111726</td>\n",
       "      <td>0.223008</td>\n",
       "      <td>0.942206</td>\n",
       "      <td>0.028609</td>\n",
       "      <td>0.013988</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.0</th>\n",
       "      <td>624.809010</td>\n",
       "      <td>519.917191</td>\n",
       "      <td>38.580242</td>\n",
       "      <td>1.051660e+14</td>\n",
       "      <td>15.414596</td>\n",
       "      <td>26.497837</td>\n",
       "      <td>77.635847</td>\n",
       "      <td>0.242579</td>\n",
       "      <td>1.008643e+11</td>\n",
       "      <td>2.247872</td>\n",
       "      <td>...</td>\n",
       "      <td>0.486326</td>\n",
       "      <td>0.252223</td>\n",
       "      <td>0.232458</td>\n",
       "      <td>0.225627</td>\n",
       "      <td>0.188395</td>\n",
       "      <td>0.112148</td>\n",
       "      <td>0.204414</td>\n",
       "      <td>0.956006</td>\n",
       "      <td>0.027877</td>\n",
       "      <td>0.019465</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5</th>\n",
       "      <td>621.395693</td>\n",
       "      <td>519.250842</td>\n",
       "      <td>424.438876</td>\n",
       "      <td>7.255248e+13</td>\n",
       "      <td>93.986369</td>\n",
       "      <td>167.042788</td>\n",
       "      <td>1408.575655</td>\n",
       "      <td>0.267845</td>\n",
       "      <td>7.306966e+08</td>\n",
       "      <td>5.109035</td>\n",
       "      <td>...</td>\n",
       "      <td>0.883195</td>\n",
       "      <td>0.247405</td>\n",
       "      <td>0.222312</td>\n",
       "      <td>0.216225</td>\n",
       "      <td>0.352361</td>\n",
       "      <td>0.112018</td>\n",
       "      <td>0.203865</td>\n",
       "      <td>0.952392</td>\n",
       "      <td>0.027390</td>\n",
       "      <td>0.030737</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 110 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                      id   timestamp   derived_0     derived_1  derived_2  \\\n",
       "technical_22                                                                \n",
       "-0.5          645.628554  519.521150   10.938387  1.838488e+13  28.881748   \n",
       " 0.0          624.809010  519.917191   38.580242  1.051660e+14  15.414596   \n",
       " 0.5          621.395693  519.250842  424.438876  7.255248e+13  93.986369   \n",
       "\n",
       "               derived_3    derived_4  fundamental_0  fundamental_1  \\\n",
       "technical_22                                                          \n",
       "-0.5           18.432194    17.259764       0.219686   6.027424e+08   \n",
       " 0.0           26.497837    77.635847       0.242579   1.008643e+11   \n",
       " 0.5          167.042788  1408.575655       0.267845   7.306966e+08   \n",
       "\n",
       "              fundamental_2    ...     technical_36  technical_37  \\\n",
       "technical_22                   ...                                  \n",
       "-0.5               0.959341    ...         0.312348      0.239122   \n",
       " 0.0               2.247872    ...         0.486326      0.252223   \n",
       " 0.5               5.109035    ...         0.883195      0.247405   \n",
       "\n",
       "              technical_38  technical_39  technical_40  technical_41  \\\n",
       "technical_22                                                           \n",
       "-0.5              0.240359      0.222952      0.139891      0.111726   \n",
       " 0.0              0.232458      0.225627      0.188395      0.112148   \n",
       " 0.5              0.222312      0.216225      0.352361      0.112018   \n",
       "\n",
       "              technical_42  technical_43  technical_44         y  \n",
       "technical_22                                                      \n",
       "-0.5              0.223008      0.942206      0.028609  0.013988  \n",
       " 0.0              0.204414      0.956006      0.027877  0.019465  \n",
       " 0.5              0.203865      0.952392      0.027390  0.030737  \n",
       "\n",
       "[3 rows x 110 columns]"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gp_t22.std()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>derived_0</th>\n",
       "      <th>derived_1</th>\n",
       "      <th>derived_2</th>\n",
       "      <th>derived_3</th>\n",
       "      <th>derived_4</th>\n",
       "      <th>fundamental_0</th>\n",
       "      <th>fundamental_1</th>\n",
       "      <th>fundamental_2</th>\n",
       "      <th>...</th>\n",
       "      <th>technical_36</th>\n",
       "      <th>technical_37</th>\n",
       "      <th>technical_38</th>\n",
       "      <th>technical_39</th>\n",
       "      <th>technical_40</th>\n",
       "      <th>technical_41</th>\n",
       "      <th>technical_42</th>\n",
       "      <th>technical_43</th>\n",
       "      <th>technical_44</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>technical_34</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>-0.5</th>\n",
       "      <td>628.173075</td>\n",
       "      <td>519.543698</td>\n",
       "      <td>387.554613</td>\n",
       "      <td>1.066250e+14</td>\n",
       "      <td>16.185400</td>\n",
       "      <td>21.135446</td>\n",
       "      <td>1298.226001</td>\n",
       "      <td>0.220321</td>\n",
       "      <td>1.007155e+11</td>\n",
       "      <td>4.057551</td>\n",
       "      <td>...</td>\n",
       "      <td>0.528906</td>\n",
       "      <td>0.273009</td>\n",
       "      <td>0.246940</td>\n",
       "      <td>0.246292</td>\n",
       "      <td>0.343941</td>\n",
       "      <td>0.111510</td>\n",
       "      <td>0.242800</td>\n",
       "      <td>0.966754</td>\n",
       "      <td>0.027496</td>\n",
       "      <td>0.022865</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.0</th>\n",
       "      <td>634.327724</td>\n",
       "      <td>519.240119</td>\n",
       "      <td>159.631020</td>\n",
       "      <td>5.894007e+13</td>\n",
       "      <td>95.622799</td>\n",
       "      <td>13.489698</td>\n",
       "      <td>488.841168</td>\n",
       "      <td>0.219669</td>\n",
       "      <td>1.390344e+07</td>\n",
       "      <td>3.831787</td>\n",
       "      <td>...</td>\n",
       "      <td>0.620577</td>\n",
       "      <td>0.241999</td>\n",
       "      <td>0.229720</td>\n",
       "      <td>0.219095</td>\n",
       "      <td>0.279228</td>\n",
       "      <td>0.112010</td>\n",
       "      <td>0.204950</td>\n",
       "      <td>0.956615</td>\n",
       "      <td>0.027873</td>\n",
       "      <td>0.022183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5</th>\n",
       "      <td>628.943363</td>\n",
       "      <td>519.856271</td>\n",
       "      <td>54.748925</td>\n",
       "      <td>4.153342e+13</td>\n",
       "      <td>20.348933</td>\n",
       "      <td>167.011531</td>\n",
       "      <td>175.239050</td>\n",
       "      <td>0.237089</td>\n",
       "      <td>4.047384e+00</td>\n",
       "      <td>0.801616</td>\n",
       "      <td>...</td>\n",
       "      <td>0.669042</td>\n",
       "      <td>0.220151</td>\n",
       "      <td>0.218515</td>\n",
       "      <td>0.196411</td>\n",
       "      <td>0.296914</td>\n",
       "      <td>0.112373</td>\n",
       "      <td>0.178496</td>\n",
       "      <td>0.952512</td>\n",
       "      <td>0.028529</td>\n",
       "      <td>0.022208</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 110 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                      id   timestamp   derived_0     derived_1  derived_2  \\\n",
       "technical_34                                                                \n",
       "-0.5          628.173075  519.543698  387.554613  1.066250e+14  16.185400   \n",
       " 0.0          634.327724  519.240119  159.631020  5.894007e+13  95.622799   \n",
       " 0.5          628.943363  519.856271   54.748925  4.153342e+13  20.348933   \n",
       "\n",
       "               derived_3    derived_4  fundamental_0  fundamental_1  \\\n",
       "technical_34                                                          \n",
       "-0.5           21.135446  1298.226001       0.220321   1.007155e+11   \n",
       " 0.0           13.489698   488.841168       0.219669   1.390344e+07   \n",
       " 0.5          167.011531   175.239050       0.237089   4.047384e+00   \n",
       "\n",
       "              fundamental_2    ...     technical_36  technical_37  \\\n",
       "technical_34                   ...                                  \n",
       "-0.5               4.057551    ...         0.528906      0.273009   \n",
       " 0.0               3.831787    ...         0.620577      0.241999   \n",
       " 0.5               0.801616    ...         0.669042      0.220151   \n",
       "\n",
       "              technical_38  technical_39  technical_40  technical_41  \\\n",
       "technical_34                                                           \n",
       "-0.5              0.246940      0.246292      0.343941      0.111510   \n",
       " 0.0              0.229720      0.219095      0.279228      0.112010   \n",
       " 0.5              0.218515      0.196411      0.296914      0.112373   \n",
       "\n",
       "              technical_42  technical_43  technical_44         y  \n",
       "technical_34                                                      \n",
       "-0.5              0.242800      0.966754      0.027496  0.022865  \n",
       " 0.0              0.204950      0.956615      0.027873  0.022183  \n",
       " 0.5              0.178496      0.952512      0.028529  0.022208  \n",
       "\n",
       "[3 rows x 110 columns]"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gp_t34.std()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Let's fill it "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>derived_0</th>\n",
       "      <th>derived_1</th>\n",
       "      <th>derived_2</th>\n",
       "      <th>derived_3</th>\n",
       "      <th>derived_4</th>\n",
       "      <th>fundamental_0</th>\n",
       "      <th>fundamental_1</th>\n",
       "      <th>fundamental_2</th>\n",
       "      <th>...</th>\n",
       "      <th>technical_36</th>\n",
       "      <th>technical_37</th>\n",
       "      <th>technical_38</th>\n",
       "      <th>technical_39</th>\n",
       "      <th>technical_40</th>\n",
       "      <th>technical_41</th>\n",
       "      <th>technical_42</th>\n",
       "      <th>technical_43</th>\n",
       "      <th>technical_44</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>0.370326</td>\n",
       "      <td>-0.006316</td>\n",
       "      <td>0.222831</td>\n",
       "      <td>-0.213030</td>\n",
       "      <td>0.729277</td>\n",
       "      <td>-0.335633</td>\n",
       "      <td>0.113292</td>\n",
       "      <td>1.621238</td>\n",
       "      <td>...</td>\n",
       "      <td>0.775208</td>\n",
       "      <td>-0.092491</td>\n",
       "      <td>-0.083389</td>\n",
       "      <td>-0.074990</td>\n",
       "      <td>-0.414776</td>\n",
       "      <td>0.005634</td>\n",
       "      <td>-0.012823</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>0.000456</td>\n",
       "      <td>-0.011753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>11</td>\n",
       "      <td>0</td>\n",
       "      <td>0.014765</td>\n",
       "      <td>-0.038064</td>\n",
       "      <td>-0.017425</td>\n",
       "      <td>0.320652</td>\n",
       "      <td>-0.034134</td>\n",
       "      <td>0.004413</td>\n",
       "      <td>0.114285</td>\n",
       "      <td>-0.210185</td>\n",
       "      <td>...</td>\n",
       "      <td>0.025590</td>\n",
       "      <td>-0.076664</td>\n",
       "      <td>-0.076388</td>\n",
       "      <td>-0.061450</td>\n",
       "      <td>-0.273607</td>\n",
       "      <td>0.004962</td>\n",
       "      <td>-0.002897</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>0.000848</td>\n",
       "      <td>-0.001240</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.010622</td>\n",
       "      <td>-0.050577</td>\n",
       "      <td>3.379575</td>\n",
       "      <td>-0.157525</td>\n",
       "      <td>-0.068550</td>\n",
       "      <td>-0.155937</td>\n",
       "      <td>1.219439</td>\n",
       "      <td>-0.764516</td>\n",
       "      <td>...</td>\n",
       "      <td>0.151881</td>\n",
       "      <td>-0.104106</td>\n",
       "      <td>-0.085064</td>\n",
       "      <td>-0.082283</td>\n",
       "      <td>-0.175710</td>\n",
       "      <td>0.005118</td>\n",
       "      <td>-0.035272</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>-0.000178</td>\n",
       "      <td>-0.020940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>25</td>\n",
       "      <td>0</td>\n",
       "      <td>0.138242</td>\n",
       "      <td>0.067212</td>\n",
       "      <td>-1.206004</td>\n",
       "      <td>-0.355480</td>\n",
       "      <td>-0.261329</td>\n",
       "      <td>0.178495</td>\n",
       "      <td>0.379278</td>\n",
       "      <td>-0.007262</td>\n",
       "      <td>...</td>\n",
       "      <td>1.035936</td>\n",
       "      <td>-0.104106</td>\n",
       "      <td>-0.085064</td>\n",
       "      <td>-0.082283</td>\n",
       "      <td>-0.211506</td>\n",
       "      <td>0.005118</td>\n",
       "      <td>-0.035272</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>-0.000178</td>\n",
       "      <td>-0.015959</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>26</td>\n",
       "      <td>0</td>\n",
       "      <td>0.176693</td>\n",
       "      <td>-0.025284</td>\n",
       "      <td>-0.057680</td>\n",
       "      <td>0.015100</td>\n",
       "      <td>0.180894</td>\n",
       "      <td>0.139445</td>\n",
       "      <td>-0.125687</td>\n",
       "      <td>-0.018707</td>\n",
       "      <td>...</td>\n",
       "      <td>0.630232</td>\n",
       "      <td>-0.076664</td>\n",
       "      <td>-0.076388</td>\n",
       "      <td>-0.061450</td>\n",
       "      <td>-0.001957</td>\n",
       "      <td>0.004962</td>\n",
       "      <td>-0.002897</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000848</td>\n",
       "      <td>-0.007338</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 111 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   id  timestamp  derived_0  derived_1  derived_2  derived_3  derived_4  \\\n",
       "0  10          0   0.370326  -0.006316   0.222831  -0.213030   0.729277   \n",
       "1  11          0   0.014765  -0.038064  -0.017425   0.320652  -0.034134   \n",
       "2  12          0  -0.010622  -0.050577   3.379575  -0.157525  -0.068550   \n",
       "3  25          0   0.138242   0.067212  -1.206004  -0.355480  -0.261329   \n",
       "4  26          0   0.176693  -0.025284  -0.057680   0.015100   0.180894   \n",
       "\n",
       "   fundamental_0  fundamental_1  fundamental_2    ...     technical_36  \\\n",
       "0      -0.335633       0.113292       1.621238    ...         0.775208   \n",
       "1       0.004413       0.114285      -0.210185    ...         0.025590   \n",
       "2      -0.155937       1.219439      -0.764516    ...         0.151881   \n",
       "3       0.178495       0.379278      -0.007262    ...         1.035936   \n",
       "4       0.139445      -0.125687      -0.018707    ...         0.630232   \n",
       "\n",
       "   technical_37  technical_38  technical_39  technical_40  technical_41  \\\n",
       "0     -0.092491     -0.083389     -0.074990     -0.414776      0.005634   \n",
       "1     -0.076664     -0.076388     -0.061450     -0.273607      0.004962   \n",
       "2     -0.104106     -0.085064     -0.082283     -0.175710      0.005118   \n",
       "3     -0.104106     -0.085064     -0.082283     -0.211506      0.005118   \n",
       "4     -0.076664     -0.076388     -0.061450     -0.001957      0.004962   \n",
       "\n",
       "   technical_42  technical_43  technical_44         y  \n",
       "0     -0.012823          -2.0      0.000456 -0.011753  \n",
       "1     -0.002897          -2.0      0.000848 -0.001240  \n",
       "2     -0.035272          -2.0     -0.000178 -0.020940  \n",
       "3     -0.035272          -2.0     -0.000178 -0.015959  \n",
       "4     -0.002897           0.0      0.000848 -0.007338  \n",
       "\n",
       "[5 rows x 111 columns]"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Creat dummy features "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "t22_dummy = pd.get_dummies(train.technical_22, prefix='t22')\n",
    "train = pd.concat([train, t22_dummy], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "t34_dummy = pd.get_dummies(train.technical_34, prefix='t34')\n",
    "train = pd.concat([train, t34_dummy], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "train.drop(['technical_22', 'technical_34'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 1710756 entries, 0 to 1710755\n",
      "Columns: 115 entries, id to t34_0.5\n",
      "dtypes: float32(107), int16(2), uint8(6)\n",
      "memory usage: 727.7 MB\n"
     ]
    }
   ],
   "source": [
    "train.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "origin_features = twosigmafunc.origin_features(train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import ExtraTreesRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "etr = ExtraTreesRegressor(n_estimators=40, max_depth=4, max_features='sqrt', min_samples_leaf=10, random_state=11, n_jobs=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_train, y_train, X_test, y_test = twosigmafunc.split_data(train, origin_features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ExtraTreesRegressor(bootstrap=False, criterion='mse', max_depth=4,\n",
       "          max_features='sqrt', max_leaf_nodes=None,\n",
       "          min_impurity_split=1e-07, min_samples_leaf=10,\n",
       "          min_samples_split=2, min_weight_fraction_leaf=0.0,\n",
       "          n_estimators=40, n_jobs=-1, oob_score=False, random_state=11,\n",
       "          verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "etr.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "pred = etr.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.010193361807766315"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "twosigmafunc.R_score(pred, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.021163241202072258"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred = etr.predict(X_train)\n",
    "twosigmafunc.R_score(pred, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>feature</th>\n",
       "      <th>important</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>fundamental_51</td>\n",
       "      <td>0.083280</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>fundamental_11</td>\n",
       "      <td>0.078693</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>technical_30</td>\n",
       "      <td>0.069579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>85</th>\n",
       "      <td>technical_20</td>\n",
       "      <td>0.043647</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>technical_17</td>\n",
       "      <td>0.037244</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101</th>\n",
       "      <td>technical_40</td>\n",
       "      <td>0.030443</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73</th>\n",
       "      <td>technical_6</td>\n",
       "      <td>0.029407</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>technical_2</td>\n",
       "      <td>0.028661</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>84</th>\n",
       "      <td>technical_19</td>\n",
       "      <td>0.028472</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>74</th>\n",
       "      <td>technical_7</td>\n",
       "      <td>0.027783</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>fundamental_50</td>\n",
       "      <td>0.023865</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>fundamental_23</td>\n",
       "      <td>0.022882</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>fundamental_53</td>\n",
       "      <td>0.022882</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>104</th>\n",
       "      <td>technical_43</td>\n",
       "      <td>0.021803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>technical_11</td>\n",
       "      <td>0.017721</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>fundamental_2</td>\n",
       "      <td>0.017230</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>fundamental_55</td>\n",
       "      <td>0.015757</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>fundamental_12</td>\n",
       "      <td>0.015524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>fundamental_0</td>\n",
       "      <td>0.014284</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>fundamental_57</td>\n",
       "      <td>0.014252</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>fundamental_21</td>\n",
       "      <td>0.013055</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>fundamental_42</td>\n",
       "      <td>0.013041</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>fundamental_32</td>\n",
       "      <td>0.012768</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>86</th>\n",
       "      <td>technical_21</td>\n",
       "      <td>0.012081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>fundamental_8</td>\n",
       "      <td>0.011765</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>technical_35</td>\n",
       "      <td>0.011715</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>t34_-0.5</td>\n",
       "      <td>0.011455</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>fundamental_46</td>\n",
       "      <td>0.010934</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>derived_2</td>\n",
       "      <td>0.010605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>fundamental_40</td>\n",
       "      <td>0.010330</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            feature  important\n",
       "55   fundamental_51   0.083280\n",
       "15   fundamental_11   0.078693\n",
       "92     technical_30   0.069579\n",
       "85     technical_20   0.043647\n",
       "82     technical_17   0.037244\n",
       "101    technical_40   0.030443\n",
       "73      technical_6   0.029407\n",
       "70      technical_2   0.028661\n",
       "84     technical_19   0.028472\n",
       "74      technical_7   0.027783\n",
       "54   fundamental_50   0.023865\n",
       "27   fundamental_23   0.022882\n",
       "57   fundamental_53   0.022882\n",
       "104    technical_43   0.021803\n",
       "77     technical_11   0.017721\n",
       "7     fundamental_2   0.017230\n",
       "59   fundamental_55   0.015757\n",
       "16   fundamental_12   0.015524\n",
       "5     fundamental_0   0.014284\n",
       "61   fundamental_57   0.014252\n",
       "25   fundamental_21   0.013055\n",
       "46   fundamental_42   0.013041\n",
       "36   fundamental_32   0.012768\n",
       "86     technical_21   0.012081\n",
       "12    fundamental_8   0.011765\n",
       "96     technical_35   0.011715\n",
       "109        t34_-0.5   0.011455\n",
       "50   fundamental_46   0.010934\n",
       "2         derived_2   0.010605\n",
       "44   fundamental_40   0.010330"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "imp = pd.DataFrame({'feature':origin_features, 'important':etr.feature_importances_}).sort_values('important', ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>feature</th>\n",
       "      <th>important</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>technical_16</td>\n",
       "      <td>0.000688</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>technical_44</td>\n",
       "      <td>0.000672</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>fundamental_36</td>\n",
       "      <td>0.000557</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>fundamental_35</td>\n",
       "      <td>0.000440</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69</th>\n",
       "      <td>technical_1</td>\n",
       "      <td>0.000065</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>fundamental_38</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>fundamental_45</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>fundamental_34</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>fundamental_47</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>fundamental_27</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>technical_41</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>fundamental_25</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>fundamental_20</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>62</th>\n",
       "      <td>fundamental_58</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>fundamental_59</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>fundamental_9</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>fundamental_63</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>fundamental_6</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>fundamental_1</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>fundamental_52</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            feature  important\n",
       "81     technical_16   0.000688\n",
       "105    technical_44   0.000672\n",
       "40   fundamental_36   0.000557\n",
       "39   fundamental_35   0.000440\n",
       "69      technical_1   0.000065\n",
       "42   fundamental_38   0.000000\n",
       "49   fundamental_45   0.000000\n",
       "38   fundamental_34   0.000000\n",
       "51   fundamental_47   0.000000\n",
       "31   fundamental_27   0.000000\n",
       "102    technical_41   0.000000\n",
       "29   fundamental_25   0.000000\n",
       "24   fundamental_20   0.000000\n",
       "62   fundamental_58   0.000000\n",
       "63   fundamental_59   0.000000\n",
       "13    fundamental_9   0.000000\n",
       "67   fundamental_63   0.000000\n",
       "10    fundamental_6   0.000000\n",
       "6     fundamental_1   0.000000\n",
       "56   fundamental_52   0.000000"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "imp[-20:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "      feature  important\n",
      "106  t22_-0.5   0.001463\n",
      "     feature  important\n",
      "107  t22_0.0   0.004299\n",
      "     feature  important\n",
      "108  t22_0.5   0.001567\n",
      "      feature  important\n",
      "109  t34_-0.5   0.011455\n",
      "     feature  important\n",
      "110  t34_0.0   0.001859\n",
      "     feature  important\n",
      "111  t34_0.5   0.003361\n"
     ]
    }
   ],
   "source": [
    "for col in list(t22_dummy.columns) + list(t34_dummy.columns):\n",
    "    print(imp.loc[imp.feature == col])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.4.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}