Assignment 1 - Regression and Classification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# CS156 Assignment 1\n",
    "\n",
    "_Yoav Rabinovich, September 2018_\n",
    "_________________________"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### *Moore's Law*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from scipy import stats\n",
    "import datetime as dt \n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.dates as mdates\n",
    "import math"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "D:\\Program Files\\Anaconda2\\lib\\site-packages\\IPython\\core\\interactiveshell.py:2717: DtypeWarning: Columns (3) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  interactivity=interactivity, compiler=compiler, result=result)\n",
      "D:\\Program Files\\Anaconda2\\lib\\site-packages\\ipykernel\\__main__.py:11: 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"
     ]
    }
   ],
   "source": [
    "#loading data\n",
    "benchmarks = pd.read_csv('benchmarks.csv')\n",
    "\n",
    "#extracting the date at datetime\n",
    "date = benchmarks['testID']\n",
    "date = date.str.extract(r'(-[0-9]+-)', expand=False)\n",
    "benchmarks['testID'] = date.str.replace('-', '')\n",
    "benchmarks = benchmarks.dropna()\n",
    "#choosing test\n",
    "test = benchmarks[benchmarks['benchName'] == '416.gamess']\n",
    "test['testID'] = test['testID'].apply(lambda x: pd.to_datetime(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#sorting by dates\n",
    "test = test.sort_values(by='testID')\n",
    "test = test.sort_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfsAAAFnCAYAAAChL+DqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xb76feb8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#training linear regression on log of data to fit exponential curve\n",
    "x = test['testID']\n",
    "#converting to ordinal for regression\n",
    "x2 = pd.to_datetime(test['testID'])\n",
    "x2 = x2.map(dt.datetime.toordinal)\n",
    "\n",
    "y = test['base']\n",
    "y2=y.apply(lambda x: math.log(x,2))\n",
    "\n",
    "slope, intercept, _, _, _ = stats.linregress(x2, y2)\n",
    "line = 2**(slope*x2+intercept)\n",
    "\n",
    "#plotting on semi-log graph, exponential fit straightens out\n",
    "plt.yscale('log',basey=2)\n",
    "plt.plot_date(x, y, \"o\", markersize=1)\n",
    "plt.plot_date(x, line, '-r')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "This analysis doesn't support Moore's law, since computing speed seems to double every ~6 years rather than 2. However, more benchmarks should be tested over a longer period of time."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Digit Classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "D:\\Program Files\\Anaconda2\\lib\\site-packages\\sklearn\\cross_validation.py:41: DeprecationWarning: This module was deprecated in version 0.18 in favor of the model_selection module into which all the refactored classes and functions are moved. Also note that the interface of the new CV iterators are different from that of this module. This module will be removed in 0.20.\n",
      "  \"This module will be removed in 0.20.\", DeprecationWarning)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.cross_validation import train_test_split\n",
    "from sklearn.neighbors import KNeighborsClassifier as KNN\n",
    "from sklearn.metrics import confusion_matrix, classification_report\n",
    "from sklearn.datasets import load_digits\n",
    "from sklearn.model_selection import GridSearchCV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#loading data\n",
    "dataset = load_digits()\n",
    "X = dataset[\"data\"]\n",
    "y = dataset[\"target\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10733390>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAVgAAAFdCAYAAABGoXXzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAAPYQAAD2EBqD+naQAAEUFJREFUeJzt3X+s3XV9x/Hnq4yIQudNxOGcssoQhsEgP/w1B90EAUnE\nkiwoEpemYaYoCbrNKHEG2R/GjEiJYM3cnEVQF2Ji1QyEIWMIyBqKt45RNEEqnYMqPwamhfGjn/1x\nTufl2mt3br/v+73n8Hwk949zrOe80pbn/fZ7zv2etNaQJHVvSd8DJGlSGVhJKmJgJamIgZWkIgZW\nkooYWEkqYmAlqYiBlaQiBlaSihhYSSoyNoFN8oEk9yV5IsntSV7f9yaAJMcn+WaSnybZmeT0vjcB\nJLkgyYYkjyfZluTrSQ7rexdAktVJNiV5bPh1W5JT+941W5KPDv9ML1kEWy4cbpn5dXffu3ZJ8vIk\nVyZ5KMmO4Z/vMYtg1327+X3bmeSyhXj+sQhskncBnwYuBI4GNgHXJTmw12ED+wPTwPuBxXRhh+OB\ny4A3AicB+wLXJ3lhr6sGtgIfAY4BjgVuBL6R5IheV80w/Ab+PgZ/1xaLu4CDgJcNv/6w3zkDSaaA\nW4H/AU4BjgD+Ani0z11Dx/HL36+XAW9j8N/p1Qvx5BmHi70kuR34t9ba+cPbYfAf6Wdaa3/T67gZ\nkuwEVrTWvtn3ltmG34x+BpzQWrul7z2zJXkY+MvW2hcXwZYDgI3AucDHge+31v68500XAu9srfV+\nVDhbkk8Bb26tLe97y54kuRQ4rbW2IP+aW/RHsEn2ZXCU851d97XBd4UbgDf3tWsMTTH4zv1I30Nm\nSrIkybuBFwHf63vP0GeBb7XWbux7yCyvHp6KujfJVUle2fegoXcAdyS5eng66s4k5/Q9arZhS84G\nvrBQz7noAwscCOwDbJt1/zYGh/zag+ER/6XALa21RXHeLsmRSX7B4J+Va4EzWmv39DyLYexfB1zQ\n95ZZbgdWMvgn+GrgVcDNSfbvc9TQIQyO9n8InAx8DvhMkvf2uupXnQG8GLhioZ7wNxbqidSrtcBr\ngLf0PWSGe4CjGPyF/xPgS0lO6DOySV7B4BvRSa21p/vasTuttetm3LwryQbgJ8CZQN+nVZYAG1pr\nHx/e3pTkSAbfCK7sb9avWAVc21p7cKGecByOYB8CnmVwcn+mg4AF+40aV0kuB04D/qi19kDfe3Zp\nrT3TWvtxa+37rbWPMXgx6fyeZx0LvBS4M8nTSZ4GlgPnJ3lq+C+BRaG19hjwI+DQvrcADwCbZ923\nGTi4hy27leRgBi/2/t1CPu+iD+zwSGIjcOKu+4Z/0U8Ebutr1zgYxvWdwB+31u7ve88eLAFe0POG\nG4DXMjhFcNTw6w7gKuCotoheER6+EHcog7j17Vbg8Fn3Hc7gCHuxWMXgtOI1C/mk43KK4BJgXZKN\nwAbgQwxeFFnX5yiA4TmwQ4FdRzeHJDkKeKS1trXHXWuBs4DTge1Jdv0L4LHW2pN97QJI8kngWuB+\nYCmDFx6WMzh/15vW2nbgOeeok2wHHm6tzT5CW1BJLga+xSBavwNcBDwNfLXPXUNrgFuTXMDg7U9v\nBM4B/qzXVUPDA7KVwLrW2s4FffLW2lh8MXif6RbgCQavNh/X96bhruXATganMWZ+/UPPu3a36Vng\nTxfB79nfAz8e/lk+CFwPvLXvXXNsvRG4ZBHs+Crwn8Pfs/uBrwCv6nvXjH2nAT8AdgD/Aazqe9OM\nbW8b/t0/dKGfeyzeBytJ42jRn4OVpHFlYCWpiIGVpCIGVpKKGFhJKmJgJalI6Q8aJHkJg4tTbAF6\nfXO7JHVkP2AZcF1r7eFf9wurf5LrFODLxc8hSX04m8EPfMypOrBbih9fPfj85z/f2WNdfvnlnHfe\neZ081tKlSzt5nF0uvvhiPvzhD3fyWGeddVYnj6NFZcuefkF1YD0tMIEOO6y7i8EfcMABnT3e1NRU\nJ4+zy9KlSzniiEXzKTZafPbYN1/kkqQiBlaSihhYSSpiYNWrE088cc+/qCennnpq3xM05gyserWY\nA/v2t7+97wkacwZWkooYWEkqYmAlqYiBlaQiBlaSihhYSSoyr8Am+UCS+5I8keT2JK/vepgkjbuR\nA5vkXcCngQuBo4FNwHVJDux4mySNtfkcwX4I+NvW2pdaa/cAq4EdwKpOl0nSmBspsEn2BY4FvrPr\nvtZaA24A3tztNEkab6MewR4I7ANsm3X/NuBlnSySpAnhuwgkqciogX0IeBY4aNb9BwEPdrJIkibE\nSIFtrT0NbAT+7xJISTK8fVu30yRpvM3nM7kuAdYl2QhsYPCughcB6zrcJUljb+TAttauHr7n9a8Z\nnBqYBk5prf2863GSNM7m9amyrbW1wNqOt0jSRPFdBJJUxMBKUhEDK0lFDKwkFTGwklTEwEpSEQMr\nSUUMrCQVMbCSVMTASlIRAytJRQysJBUxsJJUZF5X01K9FStW9D1hTsuXL+97wm5ddNFFfU+QnsMj\nWEkqYmAlqYiBlaQiBlaSihhYSSpiYCWpiIGVpCIGVpKKGFhJKmJgJamIgZWkIgZWkooYWEkqYmAl\nqcjIgU1yfJJvJvlpkp1JTq8YJknjbj5HsPsD08D7gdbtHEmaHCNfcLu19m3g2wBJ0vkiSZoQnoOV\npCIGVpKKGFhJKmJgJamIgZWkIiO/iyDJ/sChwK53EByS5Cjgkdba1i7HSdI4GzmwwHHAvzB4D2wD\nPj28/wpgVUe7JGnszed9sP+KpxYkaY8MpSQVMbCSVMTASlIRAytJRQysJBUxsJJUxMBKUhEDK0lF\nDKwkFTGwklTEwEpSEQMrSUUMrCQVmc/lCrUAPvGJT/Q9YeysX7++7wnSc3gEK0lFDKwkFTGwklTE\nwEpSEQMrSUUMrCQVMbCSVMTASlIRAytJRQysJBUxsJJUxMBKUhEDK0lFRgpskguSbEjyeJJtSb6e\n5LCqcZI0zkY9gj0euAx4I3ASsC9wfZIXdj1MksbdSNeDba2dNvN2kpXAz4BjgVu6myVJ429vz8FO\nAQ14pIMtkjRR5h3YJAEuBW5prd3d3SRJmgx785Exa4HXAG/paIskTZR5BTbJ5cBpwPGttQe6nSRJ\nk2HkwA7j+k5geWvt/u4nSdJkGCmwSdYCZwGnA9uTHDT8nx5rrT3Z9ThJGmejvsi1GvhN4Cbgv2Z8\nndntLEkaf6O+D9YfrZWk/yeDKUlFDKwkFTGwklTEwEpSEQMrSUUMrCQVMbCSVMTASlIRAytJRQys\nJBUxsJJUxMBKUhEDK0lFDKwkFdmbz+RSoampqb4nzGnTpk19T9it6enpvidIz+ERrCQVMbCSVMTA\nSlIRAytJRQysJBUxsJJUxMBKUhEDK0lFDKwkFTGwklTEwEpSEQMrSUUMrCQVGSmwSVYn2ZTkseHX\nbUlOrRonSeNs1CPYrcBHgGOAY4EbgW8kOaLrYZI07ka6Hmxr7Z9m3fVXSc4F3gRs7myVJE2AeV9w\nO8kS4EzgRcD3OlskSRNi5MAmOZJBUPcDfgGc0Vq7p+thkjTu5vMugnuAo4A3AJ8DvpTk9ztdJUkT\nYOQj2NbaM8CPhze/n+QNwPnAuV0Ok6Rx18X7YJcAL+jgcSRpoox0BJvkk8C1wP3AUuBsYDlwcvfT\nJGm8jXqK4LeAK4DfBh4DfgCc3Fq7sethkjTuRn0f7DlVQyRp0ngtAkkqYmAlqYiBlaQiBlaSihhY\nSSpiYCWpiIGVpCIGVpKKGFhJKmJgJamIgZWkIgZWkooYWEkqYmAlqci8P1VWtaampvqeMKctW7b0\nPWG3PvjBD/Y9YU7r16/ve8KcFuuf5yTwCFaSihhYSSpiYCWpiIGVpCIGVpKKGFhJKmJgJamIgZWk\nIgZWkooYWEkqYmAlqYiBlaQiBlaSiuxVYJN8NMnOJJd0NUiSJsW8A5vk9cD7gE3dzZGkyTGvwCY5\nALgKOAf4704XSdKEmO8R7GeBb7XWbuxyjCRNkpE/0SDJu4HXAcd1P0eSJsdIgU3yCuBS4KTW2tM1\nkyRpMox6BHss8FLgziQZ3rcPcEKS84AXtNZalwMlaVyNGtgbgNfOum8dsBn4lHGVpF8aKbCtte3A\n3TPvS7IdeLi1trnLYZI07rr4SS6PWiVpN0Z+F8FsrbW3djFEkiaN1yKQpCIGVpKKGFhJKmJgJamI\ngZWkIgZWkooYWEkqYmAlqYiBlaQiBlaSihhYSSpiYCWpiIGVpCJ7fTUt1diyZUvfE+a0fPnyvifs\n1tTUVN8T5rRmzZq+J8zp6KOP7nvCnKanp/uesFc8gpWkIgZWkooYWEkqYmAlqYiBlaQiBlaSihhY\nSSpiYCWpiIGVpCIGVpKKGFhJKmJgJamIgZWkIgZWkoqMFNgkFybZOevr7qpxkjTO5nM92LuAE4EM\nbz/T3RxJmhzzCewzrbWfd75EkibMfM7BvjrJT5Pcm+SqJK/sfJUkTYBRA3s7sBI4BVgNvAq4Ocn+\nHe+SpLE30imC1tp1M27elWQD8BPgTOCLXQ6TpHG3V2/Taq09BvwIOLSbOZI0OfYqsEkOYBDXB7qZ\nI0mTY9T3wV6c5IQkv5vkD4CvA08DXy1ZJ0ljbNS3ab0C+ArwEuDnwC3Am1prD3c9TJLG3agvcp1V\nNUSSJo3XIpCkIgZWkooYWEkqYmAlqYiBlaQiBlaSihhYSSpiYCWpiIGVpCIGVpKKGFhJKmJgJamI\ngZWkIvP5VFktgHXr1vU9YU5r1qzpe8Jubdmype8Jc1q2bFnfE+a0YsWKvifMaXp6uu8Je8UjWEkq\nYmAlqYiBlaQiBlaSihhYSSpiYCWpiIGVpCIGVpKKGFhJKmJgJamIgZWkIgZWkooYWEkqMnJgk7w8\nyZVJHkqyI8mmJMdUjJOkcTbS5QqTTAG3At8BTgEeAl4NPNr9NEkab6NeD/ajwP2ttXNm3PeTDvdI\n0sQY9RTBO4A7klydZFuSO5Ocs8f/lyQ9D40a2EOAc4EfAicDnwM+k+S9XQ+TpHE36imCJcCG1trH\nh7c3JTkSWA1c2ekySRpzox7BPgBsnnXfZuDgbuZI0uQYNbC3AofPuu9wfKFLkn7FqIFdA7wpyQVJ\nfi/Je4BzgMu7nyZJ422kwLbW7gDOAM4C/h34GHB+a+0fC7ZJ0lgb9UUuWmvXANcUbJGkieK1CCSp\niIGVpCIGVpKKGFhJKmJgJamIgZWkIgZWkooYWEkqYmAlqYiBlaQiBlaSihhYSSpiYCWpiIGVpCIj\nX65QC2PdunV9T5jTsmXL+p6wWytXrux7wpxuuummvifMaf369X1PmFgewUpSEQMrSUUMrCQVMbCS\nVMTASlIRAytJRQysJBUxsJJUxMBKUhEDK0lFDKwkFTGwklTEwEpSkZECm+S+JDt383VZ1UBJGlej\nXq7wOGCfGbdfC1wPXN3ZIkmaECMFtrX28MzbSd4B3Nta+26nqyRpAsz7HGySfYGzgS90N0eSJsfe\nvMh1BvBi4IqOtkjSRNmbwK4Crm2tPdjVGEmaJPP6TK4kBwMnASu6nSNJk2O+R7CrgG3ANR1ukaSJ\nMnJgkwRYCaxrre3sfJEkTYj5HMGeBLwS+GLHWyRpoox8Dra19s8894cNJEm74bUIJKmIgZWkIgZW\nkooYWEkqYmAlqYiBlaQiBlaSihhYSSpiYJ8Hnnrqqb4nzGnjxo19T5jT1772tb4nzOnmm2/ue8Ju\nPfroo31PWFQM7PPAYg7snXfe2feEOS3mwH73u4vzQ0QM7HMZWEkqYmAlqYiBlaQi8/pEgxHsV/z4\nE+uZZ57p7LFaa50+3tatWzt7rCeeeKKzx5uenu7kcXZ5/PHHO3vMe++9t5PH2WX79u2dPeaOHTs6\neRyAZ599ttPHW+T22Le01sqePcl7gC+XPYEk9efs1tpXft0vqA7sS4BTgC3Ak2VPJEkLZz9gGXBd\na+3hX/cLSwMrSc9nvsglSUUMrCQVMbCSVMTASlIRAytJRQysJBUxsJJU5H8B/K60Sklavd0AAAAA\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x6067fd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#displaying example\n",
    "plt.gray() \n",
    "plt.matshow(dataset.images[6]) \n",
    "plt.show() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, error_score='raise',\n",
       "       estimator=KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='minkowski',\n",
       "           metric_params=None, n_jobs=1, n_neighbors=5, p=2,\n",
       "           weights='uniform'),\n",
       "       fit_params=None, iid=True, n_jobs=1,\n",
       "       param_grid={'n_neighbors': [1, 3, 9]}, pre_dispatch='2*n_jobs',\n",
       "       refit=True, return_train_score='warn', scoring='accuracy',\n",
       "       verbose=0)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#optimizing a 10-digit classifier (knn) with cross-validation and fitting the model\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.25, random_state = 0)\n",
    "\n",
    "parameters_n = {'n_neighbors': [1, 3, 9]}\n",
    "clf = GridSearchCV(KNN(), parameters_n,cv=5, scoring='accuracy')\n",
    "\n",
    "clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[37  0  0  0  0  0  0  0  0  0]\n",
      " [ 0 43  0  0  0  0  0  0  0  0]\n",
      " [ 0  0 43  1  0  0  0  0  0  0]\n",
      " [ 0  0  0 45  0  0  0  0  0  0]\n",
      " [ 0  0  0  0 38  0  0  0  0  0]\n",
      " [ 0  0  0  0  0 47  0  0  0  1]\n",
      " [ 0  0  0  0  0  0 52  0  0  0]\n",
      " [ 0  0  0  0  0  0  0 48  0  0]\n",
      " [ 0  0  0  0  0  0  0  0 48  0]\n",
      " [ 0  0  0  1  0  1  0  0  0 45]] \n",
      "\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       1.00      1.00      1.00        37\n",
      "          1       1.00      1.00      1.00        43\n",
      "          2       1.00      0.98      0.99        44\n",
      "          3       0.96      1.00      0.98        45\n",
      "          4       1.00      1.00      1.00        38\n",
      "          5       0.98      0.98      0.98        48\n",
      "          6       1.00      1.00      1.00        52\n",
      "          7       1.00      1.00      1.00        48\n",
      "          8       1.00      1.00      1.00        48\n",
      "          9       0.98      0.96      0.97        47\n",
      "\n",
      "avg / total       0.99      0.99      0.99       450\n",
      "\n"
     ]
    }
   ],
   "source": [
    "#presenting results\n",
    "y_pred = clf.predict(X_test)\n",
    "cm = confusion_matrix(y_test, y_pred)\n",
    "classification_rep = classification_report(y_test, y_pred)\n",
    "\n",
    "print cm, \"\\n\\n\" ,classification_rep"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#using entire* dataset, from a csv version of MNIST\n",
    "training = pd.read_csv('mnist_train.csv')\n",
    "testing = pd.read_csv('mnist_test.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#*actually taking a large subset, so it would finish this millenioum\n",
    "training_y = training.iloc[:2000,0]\n",
    "training_x = training.iloc[:2000,1:]\n",
    "\n",
    "testing_y = testing.iloc[:500,0]\n",
    "testing_x = testing.iloc[:500,1:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, error_score='raise',\n",
       "       estimator=KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='minkowski',\n",
       "           metric_params=None, n_jobs=1, n_neighbors=5, p=2,\n",
       "           weights='uniform'),\n",
       "       fit_params=None, iid=True, n_jobs=1,\n",
       "       param_grid={'n_neighbors': [1, 3, 9]}, pre_dispatch='2*n_jobs',\n",
       "       refit=True, return_train_score='warn', scoring='accuracy',\n",
       "       verbose=0)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#optimizing and fitting as done above\n",
    "parameters_n = {'n_neighbors': [1,3,9]}\n",
    "clf2 = GridSearchCV(KNN(), parameters_n,cv=5, scoring='accuracy')\n",
    "\n",
    "clf2.fit(training_x, training_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[42  0  0  0  0  0  0  0  0  0]\n",
      " [ 0 67  0  0  0  0  0  0  0  0]\n",
      " [ 0  2 44  1  0  0  1  6  1  0]\n",
      " [ 0  1  1 35  0  5  1  2  1  0]\n",
      " [ 0  1  0  0 44  0  1  0  0  9]\n",
      " [ 0  0  0  2  0 44  2  0  1  1]\n",
      " [ 2  0  0  0  1  1 39  0  0  0]\n",
      " [ 0  0  0  0  1  0  0 44  0  3]\n",
      " [ 2  0  2  0  0  1  2  2 28  3]\n",
      " [ 0  0  0  0  4  1  0  3  2 44]] \n",
      "\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.91      1.00      0.95        42\n",
      "          1       0.94      1.00      0.97        67\n",
      "          2       0.94      0.80      0.86        55\n",
      "          3       0.92      0.76      0.83        46\n",
      "          4       0.88      0.80      0.84        55\n",
      "          5       0.85      0.88      0.86        50\n",
      "          6       0.85      0.91      0.88        43\n",
      "          7       0.77      0.92      0.84        48\n",
      "          8       0.85      0.70      0.77        40\n",
      "          9       0.73      0.81      0.77        54\n",
      "\n",
      "avg / total       0.87      0.86      0.86       500\n",
      "\n"
     ]
    }
   ],
   "source": [
    "#presenting results\n",
    "y_pred2 = clf2.predict(testing_x)\n",
    "cm = confusion_matrix(testing_y, y_pred2)\n",
    "classification_rep = classification_report(testing_y, y_pred2)\n",
    "\n",
    "print cm, \"\\n\\n\" ,classification_rep"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "*Disclaimer*: This work was done in collaboration with Michelle and Kalia."
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda root]",
   "language": "python",
   "name": "conda-root-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}