Permutation Test.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import extract\n",
    "import visualize\n",
    "from sklearn import tree\n",
    "import pydotplus\n",
    "import seaborn as sns\n",
    "import time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn import preprocessing\n",
    "from sklearn.model_selection import train_test_split, GridSearchCV\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.ensemble import RandomForestRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/mark/Math/bondproject/venv/local/lib/python2.7/site-packages/pandas/core/indexing.py:477: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  self.obj[item] = s\n",
      "/home/mark/Math/bondproject/venv/local/lib/python2.7/site-packages/pandas/core/indexing.py:297: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  self.obj[key] = _infer_fill_value(value)\n"
     ]
    }
   ],
   "source": [
    "d = extract.get_data()\n",
    "d = extract.clean_data(d)\n",
    "s0 = pd.read_csv(\"./data/price_change_data.csv\", usecols = ['d_index', 'tradetype',\n",
    "\t\t'state', 'coupon', 'maturity', 'issuetype', 'issuesource', 'rtg',\n",
    "\t\t'tradedate', 'dprice', 'holdtime'])\n",
    "d_index = s0.pop('d_index')\n",
    "s_state_features = extract.build_state_features(s0, num_states = None)\n",
    "s_other_features = extract.build_other_text_features(s0)\n",
    "s = pd.concat([s0.drop(['state', 'issuetype', 'issuesource', 'tradetype'], axis = 1), s_state_features, s_other_features], axis = 1)\n",
    "s = s.dropna()\n",
    "s = s[s.holdtime < 7] #pick out only data with holdtime < 7 days\n",
    "dprice = s.pop('dprice')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "s_train, s_test, dp_train, dp_test = train_test_split(s, dprice, test_size=0.20, random_state=180)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "clf1 = RandomForestRegressor(max_depth = 25, n_estimators = 100)\n",
    "clf2 = RandomForestRegressor(max_depth = 25, n_estimators = 100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestRegressor(bootstrap=True, criterion='mse', max_depth=25,\n",
       "           max_features='auto', max_leaf_nodes=None,\n",
       "           min_impurity_split=1e-07, min_samples_leaf=1,\n",
       "           min_samples_split=2, min_weight_fraction_leaf=0.0,\n",
       "           n_estimators=100, n_jobs=1, oob_score=False, random_state=None,\n",
       "           verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf1.fit(s_train, dp_train.values)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### Permute dp_train values for permutation test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[-1.198 -4.016 -2.41  -0.159 -2.86 ]\n",
      "[-2.533 -2.398 -3.025 -1.405  0.903]\n"
     ]
    }
   ],
   "source": [
    "dp_train_permuted = dp_train.copy()\n",
    "dp_train_permuted = dp_train_permuted.reindex(np.random.permutation(dp_train_permuted.index))\n",
    "print(dp_train_permuted.values[:5])\n",
    "print(dp_train.values[:5])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestRegressor(bootstrap=True, criterion='mse', max_depth=25,\n",
       "           max_features='auto', max_leaf_nodes=None,\n",
       "           min_impurity_split=1e-07, min_samples_leaf=1,\n",
       "           min_samples_split=2, min_weight_fraction_leaf=0.0,\n",
       "           n_estimators=100, n_jobs=1, oob_score=False, random_state=None,\n",
       "           verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf2.fit(s_train, dp_train_permuted.values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "dp_predict_1 = clf1.predict(s_test)\n",
    "dp_predict_2 = clf2.predict(s_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAcMAAAGOCAYAAADrQGhsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n7RGGAAC3RxgCANweYQgAcHv/C3kOi4AlOVVnAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f83d1619cd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cnf_matrix_1 = confusion_matrix((dp_test>0).astype(np.int), (dp_predict_1>0).astype(np.int))\n",
    "plt.figure()\n",
    "visualize.plot_confusion_matrix(cnf_matrix_1, classes=[0, 1], title='Confusion matrix, without normalization')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAcMAAAGOCAYAAADrQGhsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngNsjDAEAbo8wBAC4vf8HDA5j33zgjlgAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f83d16197d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cnf_matrix_2 = confusion_matrix((dp_test>0).astype(np.int), (dp_predict_2>0).astype(np.int))\n",
    "plt.figure()\n",
    "visualize.plot_confusion_matrix(cnf_matrix_2, classes=[0, 1], title='Confusion matrix, without normalization')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}