LiamLearn/K-fold_CV_F1_score__MATT.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 'Grouped' k-fold CV\n",
    "\n",
    "### A quick demo by Matt\n",
    "\n",
    "In cross-validating, we'd like to drop out one well at a time. `LeaveOneGroupOut` is good for this:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "training_data = pd.read_csv('../training_data.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Isolate X and y:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "X = training_data.drop(['Formation', 'Well Name', 'Depth','Facies'], axis=1).values\n",
    "y = training_data['Facies'].values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We want the well names to use as groups in the k-fold analysis, so we'll get those too:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "wells = training_data[\"Well Name\"].values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we train as normal, but `LeaveOneGroupOut` gives us the approriate indices from `X` and `y` to test against one well at a time:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     CHURCHMAN BIBLE  0.498\n",
      "      CROSS H CATTLE  0.313\n",
      "            LUKE G U  0.460\n",
      "               NEWBY  0.294\n",
      "               NOLAN  0.383\n",
      "          Recruit F9  0.662\n",
      "             SHANKLE  0.477\n",
      "           SHRIMPLIN  0.476\n"
     ]
    }
   ],
   "source": [
    "from sklearn.svm import SVC\n",
    "from sklearn.model_selection import LeaveOneGroupOut\n",
    "\n",
    "logo = LeaveOneGroupOut()\n",
    "\n",
    "for train, test in logo.split(X, y, groups=wells):\n",
    "    well_name = wells[test[0]]\n",
    "    score = SVC().fit(X[train], y[train]).score(X[test], y[test])\n",
    "    print(\"{:>20s}  {:.3f}\".format(well_name, score))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}