Abalone 3 - PCA to Forest.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here's what's going down.  Imma test various component numbers and feed them to the forest.  I'll use a testing size of 0.2 and hope I'm not overfitting, and like 25 trees, turn off OOB for now because I hate the error messages."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "#Create dataframe\n",
    "df = pd.read_csv('abalone.data')\n",
    "df.columns = ['sex','length','diameter','height','whole weight','shucked weight','viscera weight','shell weight','rings']\n",
    "\n",
    "#We have to get rid of the 'sex' column for now, PCA won't dig it.\n",
    "#df = pd.get_dummies(df, columns = ['sex'])\n",
    "df.drop('sex', axis=1, inplace=True)\n",
    "\n",
    "y = df['rings']\n",
    "X = df.drop('rings', axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#This one creates new training and testing sets each time.\n",
    "def forest1(X, y, tsize, numtrees, text=True):\n",
    "    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=tsize, random_state=0)\n",
    "    \n",
    "    model = RandomForestClassifier(n_estimators=numtrees, oob_score=False).fit(X_train,y_train)\n",
    "    score = model.score(X_test,y_test)\n",
    "\n",
    "    if text == True:\n",
    "        print('With testing size ' + str(tsize) + ' and forest size of ' + str(numtrees) +' score = '+ str(score))\n",
    "\n",
    "    return score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Creates and scores a model with given train/test sets\n",
    "def forest2(X_train, y_train, X_test, y_test):\n",
    "    model = RandomForestClassifier(n_estimators=25, oob_score=False).fit(X_train,y_train)\n",
    "    score = model.score(X_test,y_test)\n",
    "\n",
    "    return score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#scales and performs pca on prepared dataframe\n",
    "def scaled_pca(dataframe, components):\n",
    "    scaled = StandardScaler().fit_transform(dataframe)\n",
    "    pca = PCA(n_components=components, svd_solver='full')\n",
    "    x = pca.fit(scaled).transform(scaled)\n",
    "    return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAEjCAYAAAAhczZxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAUESWqeBv5Z1AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2b2b98ff908>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#explore possible variations\n",
    "#forest1(X, y, tsize, numtrees, text=True)\n",
    "\n",
    "num_components = list(range(2,9,1))\n",
    "scores = []\n",
    "count = 0\n",
    "\n",
    "#for testing component numbers from 7 to 2\n",
    "\n",
    "while count < 8:\n",
    "    for num in list(range(2,8,1)):\n",
    "        scaled = scaled_pca(X, num)\n",
    "        score = forest1(scaled, y, 0.2, 25, text=False)\n",
    "        scores.append(score)\n",
    "        count = count+1\n",
    "    else:\n",
    "        scaled = StandardScaler().fit_transform(X)\n",
    "        score = forest1(scaled, y, 0.2, 25, text=False)\n",
    "        scores.append(score)\n",
    "        break\n",
    "        \n",
    "plt.scatter(x=num_components, y=scores)\n",
    "plt.suptitle('Scores at various PCA component #')\n",
    "plt.xlabel('# components')\n",
    "plt.ylabel('score')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#Well that was unhelpful."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAEjCAYAAAAhczZxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\npyozQeUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2b2b98a5898>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#for testing PCA component numbers from 7 to 2, \n",
    "#and running each condition various times to observe\n",
    "#variation introduced from the unique testing/training sets.\n",
    "\n",
    "#Axes for plotting\n",
    "num_components = list(range(2,9,1))\n",
    "scores = []\n",
    "\n",
    "#loop control; count tests the various PCA conditions, \n",
    "#test controls sample size for each condition.\n",
    "count = 0\n",
    "test = 0\n",
    "\n",
    "while test <= 15:\n",
    "    while count < 8:\n",
    "        for num in list(range(2,8,1)):\n",
    "            scaled = scaled_pca(X, num)\n",
    "            score = forest1(scaled, y, 0.2, 25, text=False)\n",
    "            scores.append(score)\n",
    "            count = count+1\n",
    "        else:\n",
    "            scaled = StandardScaler().fit_transform(X)\n",
    "            score = forest1(scaled, y, 0.2, 25, text=False)\n",
    "            scores.append(score)\n",
    "            break\n",
    "       \n",
    "    plt.scatter(x=num_components, y=scores)\n",
    "    test = test+1 \n",
    "    count = 0\n",
    "    scores = []\n",
    "\n",
    "plt.suptitle('Scores at various PCA component #')\n",
    "plt.xlabel('# components')\n",
    "plt.ylabel('score')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Annoying, but interesting!  With a testing proportion of 0.20 and forest size of 25, X is scaled and run through PCA at various sizes.  Component #8 is not run through PCA, but still scaled.  Each component number condition was run 15 times to observe the variation produced by picking different training and testing sets.\n",
    "\n",
    "Does the variation mean there are outliers that are messing with the information?\n",
    "\n",
    "...does y need to be scaled?"
   ]
  },
  {
   "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": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}