Exercise 4/Exercise 4 MNIST.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import os\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn import svm\n",
    "from sklearn import grid_search\n",
    "from math import log\n",
    "path = os.getcwd()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Number of training samples (1 - 48000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "n = 5000"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Read dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "trainingData = pd.read_csv(os.path.join(path,'data','train.csv'), index_col=False).get_values()[:,1:]\n",
    "trainingLabels = pd.read_csv(os.path.join(path,'data','train.csv'), usecols=['label']).get_values()[:,0]\n",
    "testData = pd.read_csv(os.path.join(path,'data','test.csv')).get_values()\n",
    "testLabels = pd.read_csv(os.path.join(path,'data','out.csv'), usecols=['Label']).get_values()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Hyper-parameter optimisation using crossvalidation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 7 candidates, totalling 35 fits\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.891000 -   0.6s\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.877490 -   0.8s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.865731 -   0.8s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.869261 -   0.8s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.895582 -   0.6s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.906375 -   0.6s\n",
      "[CV] ................................ C=1e-15, score=0.904000 -   0.5s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.901198 -   0.6s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.886774 -   0.6s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.924699 -   0.5s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.894422 -   0.7s\n",
      "[CV] C=1e-14 .........................................................\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  10 tasks      | elapsed:    2.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV] ................................ C=1e-14, score=0.889222 -   0.7s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.893000 -   0.6s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.905622 -   0.5s\n",
      "[CV] ................................ C=1e-14, score=0.881764 -   0.7s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.884462 -   0.8s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.876248 -   0.7s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.874000 -   0.6s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.861723 -   0.7s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.896586 -   0.8s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.884462 -   0.7s\n",
      "[CV] ................................ C=1e-12, score=0.871000 -   0.6s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.869261 -   0.7s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.865731 -   0.6s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.892570 -   0.6s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.869261 -   0.7s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.884462 -   0.8s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.871000 -   0.7s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.865731 -   0.7s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.884462 -   0.6s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.892570 -   0.7s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.869261 -   0.7s\n",
      "[CV] ................................ C=1e-10, score=0.871000 -   0.6s\n",
      "[CV] ................................ C=1e-10, score=0.865731 -   0.7s\n",
      "[CV] ................................ C=1e-10, score=0.892570 -   0.7s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  35 out of  35 | elapsed:    6.1s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 7 candidates, totalling 35 fits\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.901000 -   0.8s\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.877756 -   0.9s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.887226 -   0.9s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.898406 -   1.0s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.926295 -   0.5s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.939000 -   0.6s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.931138 -   0.7s\n",
      "[CV] ................................ C=1e-16, score=0.900602 -   0.9s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.922846 -   0.7s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.941767 -   0.6s\n",
      "[CV] ................................ C=1e-14, score=0.916168 -   0.5s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.922311 -   0.6s\n",
      "[CV] C=1e-13 .........................................................\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  10 tasks      | elapsed:    2.3s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV] ................................ C=1e-14, score=0.923000 -   0.7s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.915832 -   0.8s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.930723 -   1.0s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.920319 -   0.9s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.910180 -   0.7s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.919000 -   0.7s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.904810 -   0.7s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.929719 -   0.7s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.920319 -   0.6s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.910180 -   0.6s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.919000 -   0.7s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.904810 -   0.8s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.920319 -   0.5s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.929719 -   0.6s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.910180 -   0.5s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.904810 -   0.7s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.919000 -   0.8s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.929719 -   0.8s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.920319 -   0.7s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.910180 -   0.6s\n",
      "[CV] ................................ C=1e-10, score=0.919000 -   0.6s\n",
      "[CV] ................................ C=1e-10, score=0.904810 -   0.6s\n",
      "[CV] ................................ C=1e-10, score=0.929719 -   0.5s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  35 out of  35 | elapsed:    6.4s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 7 candidates, totalling 35 fits\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.914000 -   0.9s\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.911355 -   0.9s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.892786 -   1.1s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.898204 -   1.2s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.936255 -   0.7s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.943114 -   0.6s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.900602 -   0.8s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.947000 -   0.7s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.942771 -   0.6s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.934870 -   0.7s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.933267 -   0.8s\n",
      "[CV] C=1e-14 .........................................................\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  10 tasks      | elapsed:    2.4s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV] ................................ C=1e-14, score=0.933134 -   0.6s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.939000 -   0.6s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.932866 -   0.6s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.948795 -   0.7s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.925299 -   0.7s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.932136 -   0.6s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.941000 -   0.6s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.942771 -   0.6s\n",
      "[CV] ................................ C=1e-13, score=0.933868 -   0.7s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.925299 -   0.6s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.932136 -   0.6s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.933868 -   0.5s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.941000 -   0.6s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.942771 -   0.6s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.925299 -   0.7s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.941000 -   0.6s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.932136 -   0.6s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.933868 -   0.6s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.942771 -   0.6s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.925299 -   0.6s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.932136 -   0.7s\n",
      "[CV] ................................ C=1e-10, score=0.941000 -   0.6s\n",
      "[CV] ................................ C=1e-10, score=0.933868 -   0.5s\n",
      "[CV] ................................ C=1e-10, score=0.942771 -   0.4s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  35 out of  35 | elapsed:    6.3s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 7 candidates, totalling 35 fits\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.888224 -   1.2s\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.885772 -   1.2s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.898000 -   1.3s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.894422 -   1.3s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.954183 -   0.7s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.954000 -   0.7s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.958084 -   0.9s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.883534 -   1.2s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.944890 -   0.7s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.945783 -   0.8s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.940239 -   0.7s\n",
      "[CV] C=1e-14 .........................................................\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  10 tasks      | elapsed:    2.9s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV] ................................ C=1e-14, score=0.941118 -   0.6s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.952000 -   0.7s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.954819 -   0.6s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.944890 -   0.7s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.935259 -   0.6s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.935130 -   0.8s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.946000 -   0.7s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.939880 -   0.8s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.952811 -   0.7s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.935259 -   0.7s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.935130 -   0.8s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.946000 -   0.7s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.939880 -   0.8s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.952811 -   0.7s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.935130 -   0.7s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.935259 -   0.8s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.946000 -   0.7s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.939880 -   0.8s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.952811 -   0.7s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.935259 -   0.7s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.935130 -   0.6s\n",
      "[CV] ................................ C=1e-10, score=0.939880 -   0.5s\n",
      "[CV] ................................ C=1e-10, score=0.946000 -   0.7s\n",
      "[CV] ................................ C=1e-10, score=0.952811 -   0.5s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  35 out of  35 | elapsed:    7.2s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 7 candidates, totalling 35 fits\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.860279 -   1.5s\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.847695 -   1.7s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.871000 -   1.7s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.870518 -   1.9s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.949102 -   0.9s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.956175 -   1.0s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.957000 -   0.9s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.858434 -   1.5s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.942886 -   1.0s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.950199 -   0.8s\n",
      "[CV] C=1e-14 .........................................................\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  10 tasks      | elapsed:    3.8s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV] ................................ C=1e-14, score=0.957086 -   0.9s\n",
      "[CV] ................................ C=1e-15, score=0.949799 -   1.2s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.963000 -   0.9s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.952906 -   0.9s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.960843 -   0.9s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.951195 -   1.1s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.949102 -   1.0s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.963000 -   0.9s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.947896 -   0.8s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.957831 -   1.1s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.951195 -   0.8s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.949102 -   0.8s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.963000 -   0.8s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.947896 -   0.9s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.957831 -   0.9s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.951195 -   1.0s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.949102 -   1.0s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.947896 -   0.8s\n",
      "[CV] ................................ C=1e-11, score=0.963000 -   0.9s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.957831 -   0.9s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.951195 -   0.8s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.963000 -   0.8s\n",
      "[CV] ................................ C=1e-10, score=0.949102 -   0.8s\n",
      "[CV] ................................ C=1e-10, score=0.957831 -   0.7s\n",
      "[CV] ................................ C=1e-10, score=0.947896 -   0.8s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  35 out of  35 | elapsed:    9.2s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 7 candidates, totalling 35 fits\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.814741 -   2.6s\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.821000 -   2.6s\n",
      "[CV] ................................ C=1e-16, score=0.797595 -   2.6s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.814371 -   2.7s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.948104 -   1.2s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.945219 -   1.3s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.954000 -   1.4s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.937876 -   1.4s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.962151 -   1.3s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.942771 -   1.5s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.806225 -   2.9s\n",
      "[CV] C=1e-14 .........................................................\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  10 tasks      | elapsed:    5.6s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV] ................................ C=1e-14, score=0.960080 -   1.2s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.967000 -   1.2s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.954910 -   1.1s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.961847 -   1.2s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.952096 -   1.1s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.957171 -   1.3s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.967000 -   1.2s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.956914 -   1.2s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.961847 -   1.1s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.957171 -   1.3s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.952096 -   1.4s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.967000 -   1.5s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.956914 -   1.1s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.961847 -   1.3s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.957171 -   1.5s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.952096 -   1.3s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.967000 -   1.3s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.956914 -   1.2s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.957171 -   1.2s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.961847 -   1.5s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.952096 -   1.4s\n",
      "[CV] ................................ C=1e-10, score=0.967000 -   1.2s\n",
      "[CV] ................................ C=1e-10, score=0.961847 -   0.9s\n",
      "[CV] ................................ C=1e-10, score=0.956914 -   1.0s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  35 out of  35 | elapsed:   13.3s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 7 candidates, totalling 35 fits\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.704409 -   3.5s\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.726096 -   3.7s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.707585 -   3.8s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.692000 -   4.4s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.925299 -   2.2s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.930140 -   2.2s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.940000 -   2.3s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.711847 -   4.2s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.917836 -   2.1s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.926707 -   2.4s\n",
      "[CV] C=1e-14 .........................................................\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  10 tasks      | elapsed:    8.5s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV] ................................ C=1e-14, score=0.964143 -   2.0s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.965070 -   1.6s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.968000 -   1.6s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.964859 -   1.6s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.956914 -   1.9s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.963147 -   1.5s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.959082 -   1.9s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.964000 -   1.6s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.956914 -   1.6s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.963855 -   1.7s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.963147 -   1.6s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.959082 -   1.4s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.964000 -   1.6s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.956914 -   1.7s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.963855 -   1.5s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.963147 -   1.6s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.959082 -   1.5s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.964000 -   1.8s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.956914 -   1.9s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.963147 -   1.7s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.963855 -   1.8s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.959082 -   1.7s\n",
      "[CV] ................................ C=1e-10, score=0.956914 -   1.1s\n",
      "[CV] ................................ C=1e-10, score=0.963855 -   1.2s\n",
      "[CV] ................................ C=1e-10, score=0.964000 -   1.7s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  35 out of  35 | elapsed:   18.3s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 7 candidates, totalling 35 fits\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.351703 -   4.6s\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.356574 -   5.0s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.363000 -   5.3s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.358283 -   5.4s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.893426 -   3.3s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.880240 -   3.7s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.896000 -   3.8s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.341365 -   5.1s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.962151 -   2.1s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.875752 -   3.3s\n",
      "[CV] C=1e-14 .........................................................\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  10 tasks      | elapsed:   11.7s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV] ................................ C=1e-14, score=0.963074 -   2.4s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.885542 -   3.1s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.966000 -   2.2s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.952906 -   2.3s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.961847 -   2.3s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.965139 -   2.7s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.966068 -   2.2s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.967000 -   2.3s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.966867 -   2.3s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.957916 -   2.9s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.965139 -   2.8s\n",
      "[CV] ................................ C=1e-12, score=0.965070 -   2.5s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.967000 -   2.2s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.957916 -   2.4s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.965863 -   2.4s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.965139 -   2.4s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.965070 -   2.5s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.967000 -   2.4s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.965863 -   2.3s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.957916 -   2.4s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.965139 -   2.2s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.965070 -   2.4s\n",
      "[CV] ................................ C=1e-10, score=0.967000 -   2.2s\n",
      "[CV] ................................ C=1e-10, score=0.957916 -   2.3s\n",
      "[CV] ................................ C=1e-10, score=0.965863 -   1.8s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  35 out of  35 | elapsed:   26.1s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 7 candidates, totalling 35 fits\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.136454 -   8.0s\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.132735 -   8.5s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.134269 -   8.6s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.134000 -   9.0s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.770916 -   6.7s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.765000 -   7.2s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.766467 -   7.6s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.136546 -   8.9s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.949203 -   3.4s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.947106 -   3.8s\n",
      "[CV] C=1e-14 .........................................................\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  10 tasks      | elapsed:   20.8s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV] ................................ C=1e-15, score=0.755511 -   7.8s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.761044 -   7.5s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.955000 -   4.8s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.930862 -   3.8s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.938755 -   4.4s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.964143 -   4.0s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.969062 -   3.8s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.968000 -   3.7s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.955912 -   3.5s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.966867 -   3.6s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.965139 -   3.8s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.967066 -   3.8s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.966000 -   3.5s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.956914 -   3.4s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.965139 -   3.5s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.967871 -   3.9s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.967066 -   4.0s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.966000 -   3.9s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.956914 -   3.9s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.967871 -   4.0s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.965139 -   3.5s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.967066 -   3.5s\n",
      "[CV] ................................ C=1e-10, score=0.956914 -   3.0s\n",
      "[CV] ................................ C=1e-10, score=0.966000 -   3.4s\n",
      "[CV] ................................ C=1e-10, score=0.967871 -   2.0s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  35 out of  35 | elapsed:   44.1s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 7 candidates, totalling 35 fits\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.112000 -  16.7s\n",
      "[CV] C=1e-16 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.111554 -  16.8s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.111222 -  16.8s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.111776 -  17.4s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.175649 -  16.1s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.180000 -  15.8s\n",
      "[CV] C=1e-15 .........................................................\n",
      "[CV] ................................ C=1e-16, score=0.111446 -  16.6s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.177291 -  17.5s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.855578 -  12.2s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.849301 -  12.0s\n",
      "[CV] C=1e-14 .........................................................\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  10 tasks      | elapsed:   46.4s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV] ................................ C=1e-15, score=0.175351 -  16.5s\n",
      "[CV] C=1e-14 .........................................................\n",
      "[CV] ................................ C=1e-15, score=0.171687 -  17.1s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.863000 -  11.9s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.961155 -   7.9s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.835671 -  12.6s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-14, score=0.849398 -  13.5s\n",
      "[CV] C=1e-13 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.960080 -   8.4s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.962000 -   7.8s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.945892 -   7.7s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-13, score=0.954819 -   7.8s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.964143 -   7.0s\n",
      "[CV] C=1e-12 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.967066 -   7.9s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.968000 -   7.7s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.956914 -   7.6s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-12, score=0.965863 -   7.4s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.965139 -   7.2s\n",
      "[CV] C=1e-11 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.967066 -   8.3s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.968000 -   8.1s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.956914 -   6.8s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-11, score=0.965863 -   7.0s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.965139 -   8.5s\n",
      "[CV] C=1e-10 .........................................................\n",
      "[CV] ................................ C=1e-10, score=0.968000 -   7.5s\n",
      "[CV] ................................ C=1e-10, score=0.956914 -   6.3s\n",
      "[CV] ................................ C=1e-10, score=0.967066 -   8.1s\n",
      "[CV] ................................ C=1e-10, score=0.965863 -   4.9s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  35 out of  35 | elapsed:  1.6min finished\n"
     ]
    }
   ],
   "source": [
    "params = []\n",
    "scores = []\n",
    "clfs = []\n",
    "pcas = []\n",
    "for n_pca in [0.05*x for x in range(10,20)]:\n",
    "    # Dimensionality Reduction\n",
    "    pcas.append(PCA(n_components=n_pca).fit(trainingData[:n,:]))\n",
    "    trainingData_pca = pcas[-1].transform(trainingData[:n,:])\n",
    "    testData_pca = pcas[-1].transform(testData)\n",
    "    \n",
    "    # Crossvalidation\n",
    "    parameters = {'C':[10**x for x in range(-16,-9)]}\n",
    "    svr = svm.SVC(kernel='poly', degree=3)\n",
    "    clfs.append(grid_search.GridSearchCV(svr, parameters, n_jobs=4, cv=5, verbose=5))\n",
    "    clfs[-1].fit(trainingData_pca, trainingLabels[:n])\n",
    "\n",
    "    # Testerror\n",
    "    score_pca = clfs[-1].score(testData_pca,testLabels)\n",
    "    bestC_pca = clfs[-1].best_params_['C']\n",
    "    \n",
    "    params.append(bestC_pca)\n",
    "    scores.append(score_pca)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZYAAAEPCAYAAABhkeIdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nZtZWTixmZtZWTixmZtZW/x/+fASGtTwajAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f2cdc3521d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.xlabel('Preserved Variance');plt.ylabel('C [10^x]')\n",
    "plt.plot([0.05*x for x in range(10,20)], [log(x,10) for x in params])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY4AAAEPCAYAAABV6CMBAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n//dCy1xN7XSNAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f2cdbf3d7f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.xlabel('Preserved Variance');plt.ylabel('Accuracy')\n",
    "plt.plot([0.05*x for x in range(10,20)], scores)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Visualise and label random sample"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "randomImg = np.array([testData[np.random.randint(testData.shape[0]-1)]])\n",
    "pca = PCA(n_components=0.8).fit(trainingData)\n",
    "randomSample = pca.transform(randomImg)\n",
    "clf = svm.SVC(kernel='poly', degree=3, C=10e-14).fit(pca.transform(trainingData), trainingLabels)\n",
    "randomLabel = clf.predict(randomSample)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Full model accuracy: 0.9835\n"
     ]
    }
   ],
   "source": [
    "print ('Full model accuracy: '+str(clf.score(pca.transform(testData),testLabels)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAEKCAYAAAAy4ujqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f2cdc04b860>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(randomImg.reshape((28,28)), interpolation=\"nearest\", cmap='gray_r')\n",
    "plt.title('Predicted Label: '+str(randomLabel));plt.show()"
   ]
  }
 ],
 "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": "3.5.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}