Intro.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import random\n",
    "import seaborn.apionly as sns\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.multiclass import OutputCodeClassifier\n",
    "from sklearn.svm import LinearSVC\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Machine learning workflow\n",
    "![alt text](figs/workflow.png)\n",
    "(image taken from [here](https://www.kaggle.com/mjbahmani/a-comprehensive-deep-learning-workflow-with-python))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Task\n",
    "In this tutorial, we go through an introductory classification problem. It is called MNIST digit classification."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Acquisition\n",
    "There is a challenge on this dataset on Kaggle [here](https://www.kaggle.com/c/digit-recognizer). You should join the challenge, download the dataset and extract it into `data/`. The dataset is less than 20MB in size. It contains 42K training images and 20K validation images. Each image contains a single handwritten digit and is 28x28 pixels large."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Loading the data\n",
    "The data comes in csv format and we use pandas package to read it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "train = pd.read_csv('data/train.csv')\n",
    "test = pd.read_csv('data/test.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's take a look at the training and test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>label</th>\n",
       "      <th>pixel0</th>\n",
       "      <th>pixel1</th>\n",
       "      <th>pixel2</th>\n",
       "      <th>pixel3</th>\n",
       "      <th>pixel4</th>\n",
       "      <th>pixel5</th>\n",
       "      <th>pixel6</th>\n",
       "      <th>pixel7</th>\n",
       "      <th>pixel8</th>\n",
       "      <th>...</th>\n",
       "      <th>pixel774</th>\n",
       "      <th>pixel775</th>\n",
       "      <th>pixel776</th>\n",
       "      <th>pixel777</th>\n",
       "      <th>pixel778</th>\n",
       "      <th>pixel779</th>\n",
       "      <th>pixel780</th>\n",
       "      <th>pixel781</th>\n",
       "      <th>pixel782</th>\n",
       "      <th>pixel783</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41970</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41971</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41972</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41973</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41974</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41975</th>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41976</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41977</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41978</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41979</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41980</th>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>27</td>\n",
       "      <td>253</td>\n",
       "      <td>110</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41981</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41982</th>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41983</th>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41984</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41985</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41986</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41987</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41988</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41989</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41990</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41991</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41992</th>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41993</th>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41994</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41995</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41996</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41997</th>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41998</th>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41999</th>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>42000 rows × 785 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       label  pixel0  pixel1  pixel2  pixel3  pixel4  pixel5  pixel6  pixel7  \\\n",
       "0          1       0       0       0       0       0       0       0       0   \n",
       "1          0       0       0       0       0       0       0       0       0   \n",
       "2          1       0       0       0       0       0       0       0       0   \n",
       "3          4       0       0       0       0       0       0       0       0   \n",
       "4          0       0       0       0       0       0       0       0       0   \n",
       "5          0       0       0       0       0       0       0       0       0   \n",
       "6          7       0       0       0       0       0       0       0       0   \n",
       "7          3       0       0       0       0       0       0       0       0   \n",
       "8          5       0       0       0       0       0       0       0       0   \n",
       "9          3       0       0       0       0       0       0       0       0   \n",
       "10         8       0       0       0       0       0       0       0       0   \n",
       "11         9       0       0       0       0       0       0       0       0   \n",
       "12         1       0       0       0       0       0       0       0       0   \n",
       "13         3       0       0       0       0       0       0       0       0   \n",
       "14         3       0       0       0       0       0       0       0       0   \n",
       "15         1       0       0       0       0       0       0       0       0   \n",
       "16         2       0       0       0       0       0       0       0       0   \n",
       "17         0       0       0       0       0       0       0       0       0   \n",
       "18         7       0       0       0       0       0       0       0       0   \n",
       "19         5       0       0       0       0       0       0       0       0   \n",
       "20         8       0       0       0       0       0       0       0       0   \n",
       "21         6       0       0       0       0       0       0       0       0   \n",
       "22         2       0       0       0       0       0       0       0       0   \n",
       "23         0       0       0       0       0       0       0       0       0   \n",
       "24         2       0       0       0       0       0       0       0       0   \n",
       "25         3       0       0       0       0       0       0       0       0   \n",
       "26         6       0       0       0       0       0       0       0       0   \n",
       "27         9       0       0       0       0       0       0       0       0   \n",
       "28         9       0       0       0       0       0       0       0       0   \n",
       "29         7       0       0       0       0       0       0       0       0   \n",
       "...      ...     ...     ...     ...     ...     ...     ...     ...     ...   \n",
       "41970      2       0       0       0       0       0       0       0       0   \n",
       "41971      3       0       0       0       0       0       0       0       0   \n",
       "41972      4       0       0       0       0       0       0       0       0   \n",
       "41973      4       0       0       0       0       0       0       0       0   \n",
       "41974      3       0       0       0       0       0       0       0       0   \n",
       "41975      9       0       0       0       0       0       0       0       0   \n",
       "41976      2       0       0       0       0       0       0       0       0   \n",
       "41977      4       0       0       0       0       0       0       0       0   \n",
       "41978      4       0       0       0       0       0       0       0       0   \n",
       "41979      4       0       0       0       0       0       0       0       0   \n",
       "41980      7       0       0       0       0       0       0       0       0   \n",
       "41981      2       0       0       0       0       0       0       0       0   \n",
       "41982      8       0       0       0       0       0       0       0       0   \n",
       "41983      7       0       0       0       0       0       0       0       0   \n",
       "41984      3       0       0       0       0       0       0       0       0   \n",
       "41985      3       0       0       0       0       0       0       0       0   \n",
       "41986      0       0       0       0       0       0       0       0       0   \n",
       "41987      5       0       0       0       0       0       0       0       0   \n",
       "41988      0       0       0       0       0       0       0       0       0   \n",
       "41989      5       0       0       0       0       0       0       0       0   \n",
       "41990      3       0       0       0       0       0       0       0       0   \n",
       "41991      1       0       0       0       0       0       0       0       0   \n",
       "41992      9       0       0       0       0       0       0       0       0   \n",
       "41993      6       0       0       0       0       0       0       0       0   \n",
       "41994      4       0       0       0       0       0       0       0       0   \n",
       "41995      0       0       0       0       0       0       0       0       0   \n",
       "41996      1       0       0       0       0       0       0       0       0   \n",
       "41997      7       0       0       0       0       0       0       0       0   \n",
       "41998      6       0       0       0       0       0       0       0       0   \n",
       "41999      9       0       0       0       0       0       0       0       0   \n",
       "\n",
       "       pixel8    ...     pixel774  pixel775  pixel776  pixel777  pixel778  \\\n",
       "0           0    ...            0         0         0         0         0   \n",
       "1           0    ...            0         0         0         0         0   \n",
       "2           0    ...            0         0         0         0         0   \n",
       "3           0    ...            0         0         0         0         0   \n",
       "4           0    ...            0         0         0         0         0   \n",
       "5           0    ...            0         0         0         0         0   \n",
       "6           0    ...            0         0         0         0         0   \n",
       "7           0    ...            0         0         0         0         0   \n",
       "8           0    ...            0         0         0         0         0   \n",
       "9           0    ...            0         0         0         0         0   \n",
       "10          0    ...            0         0         0         0         0   \n",
       "11          0    ...            0         0         0         0         0   \n",
       "12          0    ...            0         0         0         0         0   \n",
       "13          0    ...            0         0         0         0         0   \n",
       "14          0    ...            0         0         0         0         0   \n",
       "15          0    ...            0         0         0         0         0   \n",
       "16          0    ...            0         0         0         0         0   \n",
       "17          0    ...            0         0         0         0         0   \n",
       "18          0    ...            0         0         0         0         0   \n",
       "19          0    ...            0         0         0         0         0   \n",
       "20          0    ...            0         0         0         0         0   \n",
       "21          0    ...            0         0         0         0         0   \n",
       "22          0    ...            0         0         0         0         0   \n",
       "23          0    ...            0         0         0         0         0   \n",
       "24          0    ...            0         0         0         0         0   \n",
       "25          0    ...            0         0         0         0         0   \n",
       "26          0    ...            0         0         0         0         0   \n",
       "27          0    ...            0         0         0         0         0   \n",
       "28          0    ...            0         0         0         0         0   \n",
       "29          0    ...            0         0         0         0         0   \n",
       "...       ...    ...          ...       ...       ...       ...       ...   \n",
       "41970       0    ...            0         0         0         0         0   \n",
       "41971       0    ...            0         0         0         0         0   \n",
       "41972       0    ...            0         0         0         0         0   \n",
       "41973       0    ...            0         0         0         0         0   \n",
       "41974       0    ...            0         0         0         0         0   \n",
       "41975       0    ...            0         0         0         0         0   \n",
       "41976       0    ...            0         0         0         0         0   \n",
       "41977       0    ...            0         0         0         0         0   \n",
       "41978       0    ...            0         0         0         0         0   \n",
       "41979       0    ...            0         0         0         0         0   \n",
       "41980       0    ...           27       253       110         0         0   \n",
       "41981       0    ...            0         0         0         0         0   \n",
       "41982       0    ...            0         0         0         0         0   \n",
       "41983       0    ...            0         0         0         0         0   \n",
       "41984       0    ...            0         0         0         0         0   \n",
       "41985       0    ...            0         0         0         0         0   \n",
       "41986       0    ...            0         0         0         0         0   \n",
       "41987       0    ...            0         0         0         0         0   \n",
       "41988       0    ...            0         0         0         0         0   \n",
       "41989       0    ...            0         0         0         0         0   \n",
       "41990       0    ...            0         0         0         0         0   \n",
       "41991       0    ...            0         0         0         0         0   \n",
       "41992       0    ...            0         0         0         0         0   \n",
       "41993       0    ...            0         0         0         0         0   \n",
       "41994       0    ...            0         0         0         0         0   \n",
       "41995       0    ...            0         0         0         0         0   \n",
       "41996       0    ...            0         0         0         0         0   \n",
       "41997       0    ...            0         0         0         0         0   \n",
       "41998       0    ...            0         0         0         0         0   \n",
       "41999       0    ...            0         0         0         0         0   \n",
       "\n",
       "       pixel779  pixel780  pixel781  pixel782  pixel783  \n",
       "0             0         0         0         0         0  \n",
       "1             0         0         0         0         0  \n",
       "2             0         0         0         0         0  \n",
       "3             0         0         0         0         0  \n",
       "4             0         0         0         0         0  \n",
       "5             0         0         0         0         0  \n",
       "6             0         0         0         0         0  \n",
       "7             0         0         0         0         0  \n",
       "8             0         0         0         0         0  \n",
       "9             0         0         0         0         0  \n",
       "10            0         0         0         0         0  \n",
       "11            0         0         0         0         0  \n",
       "12            0         0         0         0         0  \n",
       "13            0         0         0         0         0  \n",
       "14            0         0         0         0         0  \n",
       "15            0         0         0         0         0  \n",
       "16            0         0         0         0         0  \n",
       "17            0         0         0         0         0  \n",
       "18            0         0         0         0         0  \n",
       "19            0         0         0         0         0  \n",
       "20            0         0         0         0         0  \n",
       "21            0         0         0         0         0  \n",
       "22            0         0         0         0         0  \n",
       "23            0         0         0         0         0  \n",
       "24            0         0         0         0         0  \n",
       "25            0         0         0         0         0  \n",
       "26            0         0         0         0         0  \n",
       "27            0         0         0         0         0  \n",
       "28            0         0         0         0         0  \n",
       "29            0         0         0         0         0  \n",
       "...         ...       ...       ...       ...       ...  \n",
       "41970         0         0         0         0         0  \n",
       "41971         0         0         0         0         0  \n",
       "41972         0         0         0         0         0  \n",
       "41973         0         0         0         0         0  \n",
       "41974         0         0         0         0         0  \n",
       "41975         0         0         0         0         0  \n",
       "41976         0         0         0         0         0  \n",
       "41977         0         0         0         0         0  \n",
       "41978         0         0         0         0         0  \n",
       "41979         0         0         0         0         0  \n",
       "41980         0         0         0         0         0  \n",
       "41981         0         0         0         0         0  \n",
       "41982         0         0         0         0         0  \n",
       "41983         0         0         0         0         0  \n",
       "41984         0         0         0         0         0  \n",
       "41985         0         0         0         0         0  \n",
       "41986         0         0         0         0         0  \n",
       "41987         0         0         0         0         0  \n",
       "41988         0         0         0         0         0  \n",
       "41989         0         0         0         0         0  \n",
       "41990         0         0         0         0         0  \n",
       "41991         0         0         0         0         0  \n",
       "41992         0         0         0         0         0  \n",
       "41993         0         0         0         0         0  \n",
       "41994         0         0         0         0         0  \n",
       "41995         0         0         0         0         0  \n",
       "41996         0         0         0         0         0  \n",
       "41997         0         0         0         0         0  \n",
       "41998         0         0         0         0         0  \n",
       "41999         0         0         0         0         0  \n",
       "\n",
       "[42000 rows x 785 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>pixel0</th>\n",
       "      <th>pixel1</th>\n",
       "      <th>pixel2</th>\n",
       "      <th>pixel3</th>\n",
       "      <th>pixel4</th>\n",
       "      <th>pixel5</th>\n",
       "      <th>pixel6</th>\n",
       "      <th>pixel7</th>\n",
       "      <th>pixel8</th>\n",
       "      <th>pixel9</th>\n",
       "      <th>...</th>\n",
       "      <th>pixel774</th>\n",
       "      <th>pixel775</th>\n",
       "      <th>pixel776</th>\n",
       "      <th>pixel777</th>\n",
       "      <th>pixel778</th>\n",
       "      <th>pixel779</th>\n",
       "      <th>pixel780</th>\n",
       "      <th>pixel781</th>\n",
       "      <th>pixel782</th>\n",
       "      <th>pixel783</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27970</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27971</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27972</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27973</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27974</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27975</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27976</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27977</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27978</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27979</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27980</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27981</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27982</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27983</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27984</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27985</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27986</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27987</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27988</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27989</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27990</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27991</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27992</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27993</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27994</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27995</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27996</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27997</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27998</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27999</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>28000 rows × 784 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       pixel0  pixel1  pixel2  pixel3  pixel4  pixel5  pixel6  pixel7  pixel8  \\\n",
       "0           0       0       0       0       0       0       0       0       0   \n",
       "1           0       0       0       0       0       0       0       0       0   \n",
       "2           0       0       0       0       0       0       0       0       0   \n",
       "3           0       0       0       0       0       0       0       0       0   \n",
       "4           0       0       0       0       0       0       0       0       0   \n",
       "5           0       0       0       0       0       0       0       0       0   \n",
       "6           0       0       0       0       0       0       0       0       0   \n",
       "7           0       0       0       0       0       0       0       0       0   \n",
       "8           0       0       0       0       0       0       0       0       0   \n",
       "9           0       0       0       0       0       0       0       0       0   \n",
       "10          0       0       0       0       0       0       0       0       0   \n",
       "11          0       0       0       0       0       0       0       0       0   \n",
       "12          0       0       0       0       0       0       0       0       0   \n",
       "13          0       0       0       0       0       0       0       0       0   \n",
       "14          0       0       0       0       0       0       0       0       0   \n",
       "15          0       0       0       0       0       0       0       0       0   \n",
       "16          0       0       0       0       0       0       0       0       0   \n",
       "17          0       0       0       0       0       0       0       0       0   \n",
       "18          0       0       0       0       0       0       0       0       0   \n",
       "19          0       0       0       0       0       0       0       0       0   \n",
       "20          0       0       0       0       0       0       0       0       0   \n",
       "21          0       0       0       0       0       0       0       0       0   \n",
       "22          0       0       0       0       0       0       0       0       0   \n",
       "23          0       0       0       0       0       0       0       0       0   \n",
       "24          0       0       0       0       0       0       0       0       0   \n",
       "25          0       0       0       0       0       0       0       0       0   \n",
       "26          0       0       0       0       0       0       0       0       0   \n",
       "27          0       0       0       0       0       0       0       0       0   \n",
       "28          0       0       0       0       0       0       0       0       0   \n",
       "29          0       0       0       0       0       0       0       0       0   \n",
       "...       ...     ...     ...     ...     ...     ...     ...     ...     ...   \n",
       "27970       0       0       0       0       0       0       0       0       0   \n",
       "27971       0       0       0       0       0       0       0       0       0   \n",
       "27972       0       0       0       0       0       0       0       0       0   \n",
       "27973       0       0       0       0       0       0       0       0       0   \n",
       "27974       0       0       0       0       0       0       0       0       0   \n",
       "27975       0       0       0       0       0       0       0       0       0   \n",
       "27976       0       0       0       0       0       0       0       0       0   \n",
       "27977       0       0       0       0       0       0       0       0       0   \n",
       "27978       0       0       0       0       0       0       0       0       0   \n",
       "27979       0       0       0       0       0       0       0       0       0   \n",
       "27980       0       0       0       0       0       0       0       0       0   \n",
       "27981       0       0       0       0       0       0       0       0       0   \n",
       "27982       0       0       0       0       0       0       0       0       0   \n",
       "27983       0       0       0       0       0       0       0       0       0   \n",
       "27984       0       0       0       0       0       0       0       0       0   \n",
       "27985       0       0       0       0       0       0       0       0       0   \n",
       "27986       0       0       0       0       0       0       0       0       0   \n",
       "27987       0       0       0       0       0       0       0       0       0   \n",
       "27988       0       0       0       0       0       0       0       0       0   \n",
       "27989       0       0       0       0       0       0       0       0       0   \n",
       "27990       0       0       0       0       0       0       0       0       0   \n",
       "27991       0       0       0       0       0       0       0       0       0   \n",
       "27992       0       0       0       0       0       0       0       0       0   \n",
       "27993       0       0       0       0       0       0       0       0       0   \n",
       "27994       0       0       0       0       0       0       0       0       0   \n",
       "27995       0       0       0       0       0       0       0       0       0   \n",
       "27996       0       0       0       0       0       0       0       0       0   \n",
       "27997       0       0       0       0       0       0       0       0       0   \n",
       "27998       0       0       0       0       0       0       0       0       0   \n",
       "27999       0       0       0       0       0       0       0       0       0   \n",
       "\n",
       "       pixel9    ...     pixel774  pixel775  pixel776  pixel777  pixel778  \\\n",
       "0           0    ...            0         0         0         0         0   \n",
       "1           0    ...            0         0         0         0         0   \n",
       "2           0    ...            0         0         0         0         0   \n",
       "3           0    ...            0         0         0         0         0   \n",
       "4           0    ...            0         0         0         0         0   \n",
       "5           0    ...            0         0         0         0         0   \n",
       "6           0    ...            0         0         0         0         0   \n",
       "7           0    ...            0         0         0         0         0   \n",
       "8           0    ...            0         0         0         0         0   \n",
       "9           0    ...            0         0         0         0         0   \n",
       "10          0    ...            0         0         0         0         0   \n",
       "11          0    ...            0         0         0         0         0   \n",
       "12          0    ...            0         0         0         0         0   \n",
       "13          0    ...            0         0         0         0         0   \n",
       "14          0    ...            0         0         0         0         0   \n",
       "15          0    ...            0         0         0         0         0   \n",
       "16          0    ...            0         0         0         0         0   \n",
       "17          0    ...            0         0         0         0         0   \n",
       "18          0    ...            0         0         0         0         0   \n",
       "19          0    ...            0         0         0         0         0   \n",
       "20          0    ...            0         0         0         0         0   \n",
       "21          0    ...            0         0         0         0         0   \n",
       "22          0    ...            0         0         0         0         0   \n",
       "23          0    ...            0         0         0         0         0   \n",
       "24          0    ...            0         0         0         0         0   \n",
       "25          0    ...            0         0         0         0         0   \n",
       "26          0    ...            0         0         0         0         0   \n",
       "27          0    ...            0         0         0         0         0   \n",
       "28          0    ...            0         0         0         0         0   \n",
       "29          0    ...            0         0         0         0         0   \n",
       "...       ...    ...          ...       ...       ...       ...       ...   \n",
       "27970       0    ...            0         0         0         0         0   \n",
       "27971       0    ...            0         0         0         0         0   \n",
       "27972       0    ...            0         0         0         0         0   \n",
       "27973       0    ...            0         0         0         0         0   \n",
       "27974       0    ...            0         0         0         0         0   \n",
       "27975       0    ...            0         0         0         0         0   \n",
       "27976       0    ...            0         0         0         0         0   \n",
       "27977       0    ...            0         0         0         0         0   \n",
       "27978       0    ...            0         0         0         0         0   \n",
       "27979       0    ...            0         0         0         0         0   \n",
       "27980       0    ...            0         0         0         0         0   \n",
       "27981       0    ...            0         0         0         0         0   \n",
       "27982       0    ...            0         0         0         0         0   \n",
       "27983       0    ...            0         0         0         0         0   \n",
       "27984       0    ...            0         0         0         0         0   \n",
       "27985       0    ...            0         0         0         0         0   \n",
       "27986       0    ...            0         0         0         0         0   \n",
       "27987       0    ...            0         0         0         0         0   \n",
       "27988       0    ...            0         0         0         0         0   \n",
       "27989       0    ...            0         0         0         0         0   \n",
       "27990       0    ...            0         0         0         0         0   \n",
       "27991       0    ...            0         0         0         0         0   \n",
       "27992       0    ...            0         0         0         0         0   \n",
       "27993       0    ...            0         0         0         0         0   \n",
       "27994       0    ...            0         0         0         0         0   \n",
       "27995       0    ...            0         0         0         0         0   \n",
       "27996       0    ...            0         0         0         0         0   \n",
       "27997       0    ...            0         0         0         0         0   \n",
       "27998       0    ...            0         0         0         0         0   \n",
       "27999       0    ...            0         0         0         0         0   \n",
       "\n",
       "       pixel779  pixel780  pixel781  pixel782  pixel783  \n",
       "0             0         0         0         0         0  \n",
       "1             0         0         0         0         0  \n",
       "2             0         0         0         0         0  \n",
       "3             0         0         0         0         0  \n",
       "4             0         0         0         0         0  \n",
       "5             0         0         0         0         0  \n",
       "6             0         0         0         0         0  \n",
       "7             0         0         0         0         0  \n",
       "8             0         0         0         0         0  \n",
       "9             0         0         0         0         0  \n",
       "10            0         0         0         0         0  \n",
       "11            0         0         0         0         0  \n",
       "12            0         0         0         0         0  \n",
       "13            0         0         0         0         0  \n",
       "14            0         0         0         0         0  \n",
       "15            0         0         0         0         0  \n",
       "16            0         0         0         0         0  \n",
       "17            0         0         0         0         0  \n",
       "18            0         0         0         0         0  \n",
       "19            0         0         0         0         0  \n",
       "20            0         0         0         0         0  \n",
       "21            0         0         0         0         0  \n",
       "22            0         0         0         0         0  \n",
       "23            0         0         0         0         0  \n",
       "24            0         0         0         0         0  \n",
       "25            0         0         0         0         0  \n",
       "26            0         0         0         0         0  \n",
       "27            0         0         0         0         0  \n",
       "28            0         0         0         0         0  \n",
       "29            0         0         0         0         0  \n",
       "...         ...       ...       ...       ...       ...  \n",
       "27970         0         0         0         0         0  \n",
       "27971         0         0         0         0         0  \n",
       "27972         0         0         0         0         0  \n",
       "27973         0         0         0         0         0  \n",
       "27974         0         0         0         0         0  \n",
       "27975         0         0         0         0         0  \n",
       "27976         0         0         0         0         0  \n",
       "27977         0         0         0         0         0  \n",
       "27978         0         0         0         0         0  \n",
       "27979         0         0         0         0         0  \n",
       "27980         0         0         0         0         0  \n",
       "27981         0         0         0         0         0  \n",
       "27982         0         0         0         0         0  \n",
       "27983         0         0         0         0         0  \n",
       "27984         0         0         0         0         0  \n",
       "27985         0         0         0         0         0  \n",
       "27986         0         0         0         0         0  \n",
       "27987         0         0         0         0         0  \n",
       "27988         0         0         0         0         0  \n",
       "27989         0         0         0         0         0  \n",
       "27990         0         0         0         0         0  \n",
       "27991         0         0         0         0         0  \n",
       "27992         0         0         0         0         0  \n",
       "27993         0         0         0         0         0  \n",
       "27994         0         0         0         0         0  \n",
       "27995         0         0         0         0         0  \n",
       "27996         0         0         0         0         0  \n",
       "27997         0         0         0         0         0  \n",
       "27998         0         0         0         0         0  \n",
       "27999         0         0         0         0         0  \n",
       "\n",
       "[28000 rows x 784 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "You can sample a subset of the table by `sample` function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>label</th>\n",
       "      <th>pixel0</th>\n",
       "      <th>pixel1</th>\n",
       "      <th>pixel2</th>\n",
       "      <th>pixel3</th>\n",
       "      <th>pixel4</th>\n",
       "      <th>pixel5</th>\n",
       "      <th>pixel6</th>\n",
       "      <th>pixel7</th>\n",
       "      <th>pixel8</th>\n",
       "      <th>...</th>\n",
       "      <th>pixel774</th>\n",
       "      <th>pixel775</th>\n",
       "      <th>pixel776</th>\n",
       "      <th>pixel777</th>\n",
       "      <th>pixel778</th>\n",
       "      <th>pixel779</th>\n",
       "      <th>pixel780</th>\n",
       "      <th>pixel781</th>\n",
       "      <th>pixel782</th>\n",
       "      <th>pixel783</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>15501</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34833</th>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15067</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16276</th>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40920</th>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4541</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25519</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8548</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24971</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8325</th>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 785 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       label  pixel0  pixel1  pixel2  pixel3  pixel4  pixel5  pixel6  pixel7  \\\n",
       "15501      1       0       0       0       0       0       0       0       0   \n",
       "34833      6       0       0       0       0       0       0       0       0   \n",
       "15067      5       0       0       0       0       0       0       0       0   \n",
       "16276      8       0       0       0       0       0       0       0       0   \n",
       "40920      7       0       0       0       0       0       0       0       0   \n",
       "4541       2       0       0       0       0       0       0       0       0   \n",
       "25519      1       0       0       0       0       0       0       0       0   \n",
       "8548       0       0       0       0       0       0       0       0       0   \n",
       "24971      2       0       0       0       0       0       0       0       0   \n",
       "8325       6       0       0       0       0       0       0       0       0   \n",
       "\n",
       "       pixel8    ...     pixel774  pixel775  pixel776  pixel777  pixel778  \\\n",
       "15501       0    ...            0         0         0         0         0   \n",
       "34833       0    ...            0         0         0         0         0   \n",
       "15067       0    ...            0         0         0         0         0   \n",
       "16276       0    ...            0         0         0         0         0   \n",
       "40920       0    ...            0         0         0         0         0   \n",
       "4541        0    ...            0         0         0         0         0   \n",
       "25519       0    ...            0         0         0         0         0   \n",
       "8548        0    ...            0         0         0         0         0   \n",
       "24971       0    ...            0         0         0         0         0   \n",
       "8325        0    ...            0         0         0         0         0   \n",
       "\n",
       "       pixel779  pixel780  pixel781  pixel782  pixel783  \n",
       "15501         0         0         0         0         0  \n",
       "34833         0         0         0         0         0  \n",
       "15067         0         0         0         0         0  \n",
       "16276         0         0         0         0         0  \n",
       "40920         0         0         0         0         0  \n",
       "4541          0         0         0         0         0  \n",
       "25519         0         0         0         0         0  \n",
       "8548          0         0         0         0         0  \n",
       "24971         0         0         0         0         0  \n",
       "8325          0         0         0         0         0  \n",
       "\n",
       "[10 rows x 785 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.sample(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train shape = (42000, 785), test shape = (28000, 784)\n"
     ]
    }
   ],
   "source": [
    "print('train shape = {}, test shape = {}'.format(train.shape, test.shape))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's see what column names are there"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['label', 'pixel0', 'pixel1', 'pixel2', 'pixel3', 'pixel4', 'pixel5',\n",
       "       'pixel6', 'pixel7', 'pixel8',\n",
       "       ...\n",
       "       'pixel774', 'pixel775', 'pixel776', 'pixel777', 'pixel778', 'pixel779',\n",
       "       'pixel780', 'pixel781', 'pixel782', 'pixel783'],\n",
       "      dtype='object', length=785)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- What are the feature?\n",
    "- What is their type/range?\n",
    "- What is the output represented?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Is your data clean?\n",
    "Generally, you might need to look for outliers or errors in the data. Data might have been entered incorrectly or measured incorrectly. Based on the dataset and the data type, you might want to look for different kinds of errors. Some of the most obvious erros could be:\n",
    "- Missing data\n",
    "- Data with inconsistent type e.g. string or negative number as pixel value\n",
    "- Data with value extremely far away from the rest of the data\n",
    "- Noisy data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see some statstics of the data columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>label</th>\n",
       "      <th>pixel0</th>\n",
       "      <th>pixel1</th>\n",
       "      <th>pixel2</th>\n",
       "      <th>pixel3</th>\n",
       "      <th>pixel4</th>\n",
       "      <th>pixel5</th>\n",
       "      <th>pixel6</th>\n",
       "      <th>pixel7</th>\n",
       "      <th>pixel8</th>\n",
       "      <th>...</th>\n",
       "      <th>pixel774</th>\n",
       "      <th>pixel775</th>\n",
       "      <th>pixel776</th>\n",
       "      <th>pixel777</th>\n",
       "      <th>pixel778</th>\n",
       "      <th>pixel779</th>\n",
       "      <th>pixel780</th>\n",
       "      <th>pixel781</th>\n",
       "      <th>pixel782</th>\n",
       "      <th>pixel783</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>42000.000000</td>\n",
       "      <td>42000.0</td>\n",
       "      <td>42000.0</td>\n",
       "      <td>42000.0</td>\n",
       "      <td>42000.0</td>\n",
       "      <td>42000.0</td>\n",
       "      <td>42000.0</td>\n",
       "      <td>42000.0</td>\n",
       "      <td>42000.0</td>\n",
       "      <td>42000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>42000.000000</td>\n",
       "      <td>42000.000000</td>\n",
       "      <td>42000.000000</td>\n",
       "      <td>42000.00000</td>\n",
       "      <td>42000.000000</td>\n",
       "      <td>42000.000000</td>\n",
       "      <td>42000.0</td>\n",
       "      <td>42000.0</td>\n",
       "      <td>42000.0</td>\n",
       "      <td>42000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>4.456643</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.219286</td>\n",
       "      <td>0.117095</td>\n",
       "      <td>0.059024</td>\n",
       "      <td>0.02019</td>\n",
       "      <td>0.017238</td>\n",
       "      <td>0.002857</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>2.887730</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>6.312890</td>\n",
       "      <td>4.633819</td>\n",
       "      <td>3.274488</td>\n",
       "      <td>1.75987</td>\n",
       "      <td>1.894498</td>\n",
       "      <td>0.414264</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>9.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>254.000000</td>\n",
       "      <td>254.000000</td>\n",
       "      <td>253.000000</td>\n",
       "      <td>253.00000</td>\n",
       "      <td>254.000000</td>\n",
       "      <td>62.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 785 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              label   pixel0   pixel1   pixel2   pixel3   pixel4   pixel5  \\\n",
       "count  42000.000000  42000.0  42000.0  42000.0  42000.0  42000.0  42000.0   \n",
       "mean       4.456643      0.0      0.0      0.0      0.0      0.0      0.0   \n",
       "std        2.887730      0.0      0.0      0.0      0.0      0.0      0.0   \n",
       "min        0.000000      0.0      0.0      0.0      0.0      0.0      0.0   \n",
       "25%        2.000000      0.0      0.0      0.0      0.0      0.0      0.0   \n",
       "50%        4.000000      0.0      0.0      0.0      0.0      0.0      0.0   \n",
       "75%        7.000000      0.0      0.0      0.0      0.0      0.0      0.0   \n",
       "max        9.000000      0.0      0.0      0.0      0.0      0.0      0.0   \n",
       "\n",
       "        pixel6   pixel7   pixel8    ...         pixel774      pixel775  \\\n",
       "count  42000.0  42000.0  42000.0    ...     42000.000000  42000.000000   \n",
       "mean       0.0      0.0      0.0    ...         0.219286      0.117095   \n",
       "std        0.0      0.0      0.0    ...         6.312890      4.633819   \n",
       "min        0.0      0.0      0.0    ...         0.000000      0.000000   \n",
       "25%        0.0      0.0      0.0    ...         0.000000      0.000000   \n",
       "50%        0.0      0.0      0.0    ...         0.000000      0.000000   \n",
       "75%        0.0      0.0      0.0    ...         0.000000      0.000000   \n",
       "max        0.0      0.0      0.0    ...       254.000000    254.000000   \n",
       "\n",
       "           pixel776     pixel777      pixel778      pixel779  pixel780  \\\n",
       "count  42000.000000  42000.00000  42000.000000  42000.000000   42000.0   \n",
       "mean       0.059024      0.02019      0.017238      0.002857       0.0   \n",
       "std        3.274488      1.75987      1.894498      0.414264       0.0   \n",
       "min        0.000000      0.00000      0.000000      0.000000       0.0   \n",
       "25%        0.000000      0.00000      0.000000      0.000000       0.0   \n",
       "50%        0.000000      0.00000      0.000000      0.000000       0.0   \n",
       "75%        0.000000      0.00000      0.000000      0.000000       0.0   \n",
       "max      253.000000    253.00000    254.000000     62.000000       0.0   \n",
       "\n",
       "       pixel781  pixel782  pixel783  \n",
       "count   42000.0   42000.0   42000.0  \n",
       "mean        0.0       0.0       0.0  \n",
       "std         0.0       0.0       0.0  \n",
       "min         0.0       0.0       0.0  \n",
       "25%         0.0       0.0       0.0  \n",
       "50%         0.0       0.0       0.0  \n",
       "75%         0.0       0.0       0.0  \n",
       "max         0.0       0.0       0.0  \n",
       "\n",
       "[8 rows x 785 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As you saw in the data tables, for training data, the first column is the true label for the image and the rest of the columns show the value of each of 784 (= 28x28) pixels. And for test data, there is just pixel values (the label of images will be predicted by you)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# From tables to features and labels\n",
    "\n",
    "In this dataset, there is no errors in the data and we can move on. Let's split the table into features and labels for each image."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train shape: (42000, 784), Y_train sahpe: (42000,), X_test shape: (28000, 784)\n",
      "X_train type: <class 'pandas.core.frame.DataFrame'>\n"
     ]
    }
   ],
   "source": [
    "Y_train = train[\"label\"]\n",
    "X_train = train.drop(labels=\"label\",axis=1)\n",
    "X_test = test\n",
    "print('X_train shape: {}, Y_train sahpe: {}, X_test shape: {}'.format(X_train.shape, Y_train.shape, X_test.shape))\n",
    "print('X_train type: {}'.format(type(X_train)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will now reshape the pixels to their original 28x28 format (we also change the data format to numpy array)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train shape: (42000, 28, 28), Y_train sahpe: (42000,), X_test shape: (28000, 28, 28)\n",
      "X_train type: <class 'numpy.ndarray'>\n"
     ]
    }
   ],
   "source": [
    "X_train, Y_train = X_train.values.reshape(-1, 28,28), Y_train.values\n",
    "X_test = X_test.values.reshape(-1,28,28)\n",
    "# See how they are shaped now\n",
    "print('X_train shape: {}, Y_train sahpe: {}, X_test shape: {}'.format(X_train.shape, Y_train.shape, X_test.shape))\n",
    "print('X_train type: {}'.format(type(X_train)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Visualizing the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# A function for plotting images in a grid\n",
    "def plotImages(images, labels=None, nrows=2, ncols=3, shuffle=False):\n",
    "    i = 0\n",
    "    fig, ax = plt.subplots(nrows,ncols,sharex=True,sharey=True)\n",
    "    for row in range(nrows):\n",
    "        for col in range(ncols):\n",
    "            idx = i\n",
    "            if shuffle:\n",
    "                idx = random.randint(0, len(images)-1)\n",
    "            elif i >= len(images):\n",
    "                break\n",
    "            ax[row,col].imshow(images[idx], cmap='gray')\n",
    "            if labels is not None:\n",
    "                ax[row,col].set_title(\"True label :{}\".format(labels[idx]))\n",
    "            i += 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see a few images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAEICAYAAABPgw/pAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nhDp0EZFAqEMXEQmEOnQRkUD8P9GMhHTDnIAyAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x109add898>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plotImages(X_train, Y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And the frequency of the labels in the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAD8CAYAAAB+UHOxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x109add5f8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(Y_train);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Or in a more fancy way"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/saeidnaderi/anaconda3/lib/python3.6/site-packages/seaborn/categorical.py:1428: FutureWarning: remove_na is deprecated and is a private function. Do not use.\n",
      "  stat_data = remove_na(group_data)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY4AAAD8CAYAAABgmUMCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x111801d30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.countplot(Y_train);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data pre-processing\n",
    "Based on what model you are going to use, you might need to process your data in some way. Here are some examples:\n",
    "\n",
    "- If you are using SVM, you need some kind of feature for your samples. Raw pixel values is not useful there (maybe PCA is helpful)\n",
    "- Does your model put any specific constraint on its inputs? e.g. for some models such as most of neural networks, the input should be in [0,1] and you need to normalize your features\n",
    "- There might be unuseful or redundant features in the dataset and removing them is quite useful. This is called feature selection."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PCA(copy=True, iterated_power='auto', n_components=2, random_state=None,\n",
       "  svd_solver='auto', tol=0.0, whiten=False)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pca = PCA(n_components=2)\n",
    "pca.fit(X_train.reshape(-1, 784))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x116a87b00>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAD8CAYAAABkbJM/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nwrCxsbGx6RH/H2yvqwTq2PFxAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x116211668>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_PCA = pca.transform(X_train.reshape(-1, 784))\n",
    "colors = ['tab:blue', 'tab:orange', 'tab:green', 'tab:red', 'tab:purple', 'tab:brown', 'tab:pink', 'tab:gray', 'tab:olive', 'tab:cyan']\n",
    "plt.scatter(X_PCA[:,0], X_PCA[:, 1], color=[colors[i] for i in Y_train])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Apply learning\n",
    "Now that we have the data is clean and ready, it's time to think about what model to use and how to train it.\n",
    "\n",
    "A simple model for classification task is k-nearest neighbors (KNN). This model classifies a test point x based as the mode of k nearest neighbors of x among the training points. So, for training the model, you just store all the training points in the memory but, for training, you need to look into all training points to find the nearest neighbors. Check out [here](http://vision.stanford.edu/teaching/cs231n-demos/knn/) for a simple demo of KNN."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Validation set and testing the model\n",
    "**How well good is the trained model?**\n",
    "Since we do not have true labels for test data, how do we know how good the model will perform on test data? (Note that even if we had the labels, they shouldn't affect the training process in any way. Don't do it! It's cheating!) \"Validation set\" solves this problem for us. It is nothing but a small portion of training set which we put aside and don't look at while training our model and training will be done on the rest of it.\n",
    "\n",
    "**Hyper-parameter:**\n",
    "A question that comes to mind is that what \"k\" should be? It is called a \"hyper-parameter\" and most of the models have some hyper-parameters. Another place where we need validation sets is here, to choose hyper=parameters. We train the model with a specific hyper-parameter and then test the performance on the validation set and if it was not satisfactory, we change the hyper-parameter and repeat the process.\n",
    "\n",
    "**Cross-validation**\n",
    "A more effective way of testing the model performance is cross-validation. For more information, [here](https://www.cs.cmu.edu/~schneide/tut5/node42.html) and/or whatch [this](https://www.youtube.com/watch?v=sFO2ff-gTh0)\n",
    "\n",
    "Now that we need a validation set, let's provide ourselves one."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Randomly permute the dataset\n",
    "perm = np.random.permutation(len(X_train))\n",
    "X_train, Y_train = X_train[perm], Y_train[perm]\n",
    "\n",
    "# Put 2K training samples aside as the validation set remove them from training set\n",
    "X_valid, Y_valid = X_train[-2000:], Y_train[-2000:]\n",
    "X_train, Y_train = X_train[:-2000], Y_train[:-2000]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We use scikit-learn for fitting a KNN model to our dataset. scikit-learn is a machine learning package containing a wide variety of models and tools. Check out the list of supervised learning models and tools it has [here](http://scikit-learn.org/stable/supervised_learning.html)\n",
    "\n",
    "Let's train the model and predict validation set labels."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "neigh = KNeighborsClassifier(n_neighbors=10, n_jobs=-1) # n_jobs sets the number of parallel jobs. -1 means max possible.\n",
    "# KNN expects a feature vector for each sample. We reshape the training data back to the flattened form\n",
    "neigh.fit(X_train.reshape(-1, 784), Y_train)\n",
    "Y_pred = neigh.predict(X_valid.reshape(-1, 784))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see a few random validation samples and the predicted labels."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAEICAYAAABPgw/pAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3Xv8VVP+P/DXq5Smi65f+UgXl4Y0jGsYtxLGD4Nmhp9C\nMZHLMMzEQ8aQy1fF4/GTMB40oghjHpOhUCS55DYVuZSoGUKX6eJSkan0/v1xdtte2+ecs8/5nOs6\nr+fj0aP3Omufvd61Pq32WWfttWlmEBGR6teo3AmIiEhhaEAXEfGEBnQREU9oQBcR8YQGdBERT2hA\nFxHxhAb0NEhOJHldwmNnkTw7z3byfq/kTv3qJ/VrSsUO6CTXR35tIbkhUj6j3PmVG8kXSVbdTQTq\n1x9iykiSy0h+SXImyR7lzisX6tfMSvXvdZtiN5AvM2u5NSb5MYBzzey5dMeT3MbMNpcit3IjOQgA\ny51HPtSv9eoP4EwAhwL4FMAIABMA9CpnUrlQv6ZXyn+vFXuFng3J/yX5KMlHSK4DcGb8YxfJo4Mf\nrq3lnUj+g+Qqkh+R/G3CttqTfDp43xckp5DsFDusO8k5JL8K2mgbef+hJF8Prr7mkTyiAX/utgCu\nBjAs33NUshrt150BvGxmHwWD3EMAeuZ5ropUo/1a8n+vVTugB/oBeBhAawCPZjqQZCMATwKYDaAT\ngGMAXEGyb4J2GgH4C4AuALoC2ARgTOyYgcGvHZH633h00G5nAJMBDAfQDqmOfYxk+2yNkjyS5OrY\ny6MA3AFgZYK8q1Wt9esjAHYnuRvJpgAGAZiaIP9qU2v9CpT432u1D+izzGyKmW0xsw1Zjj0EwHZm\nNsLMNprZYgDjAJyerREzW2Vm/zCzDWa2FqmPxEfGDptgZgvM7GsA1wI4nSSR+qGZbGbPBHlOA/A2\ngOMStPuimXXYWiZ5EIADAdyV7b1Vrqb6FcBSAK8CWATgGwAnAxia7TxVqKb6tRz/Xit2Dj2hT3M4\ntiuALiS/jLzWGMAL2d5IsiWA2wAcC6BN8HKrDLksAbAtUv/DdwXQn2S/SH0TANNyyH3rFctdAC4x\ns+9SP3veqpl+DVwPYF+krkRXAjgbwPMke5rZt3mcr1LVTL+W699rtQ/o8W+NvwbQPFLeIRJ/CmCR\nmeWzeuAKpOY5e5nZCpIHIPVRMKpzJO4C4L8APg/avd/MLsyj3ah2APYBMCn44WgMACRXAPilmb3a\nwPNXklrqVyDVr4+Y2bKgfC/J2wDsAWBeAc5fKWqpX8vy77Xap1zi5gE4gWRbknUAfhepew3ARpJD\nSTYj2ZjkXiT3T3DeVkh9FP4imEu7tp5jBpLcg2QLpK64/mapvYkfBNCP5DFBm81I9iG5Y45/tjVI\nXcHtE/z6RfD6PgDm5HiuauNzvwKpweb/ktyeZCOS5wSv/zuPc1UTn/u1LP9efRvQxwN4H6mPUNMA\n/HVrRbB64HikloJ9DGA1gHsAbJfgvLci9UXOGqTmOuv7wupBABMBLEfqf+PLgnY/RurLoGsArALw\nCVLzo1n/7kn23vqR01JWbP0V5I+gvDHBn6GajYen/RoYAWA+UnO1XwK4GKmruLUJ/gzVbDw87ddy\n/XulHnAhIuIH367QRURqlgZ0ERFPaEAXEfFEgwZ0kseR/IDkYpJe3oouIlIt8v5SlGRjAB8idUvu\nZ0gtvepvZgsyvEffwFYIMyvYnQ7q18qhfvXWajP7n2wHNeQKvReAxWb272AZzl+RumVZREQKa0mS\ngxoyoHeCe/vsZ8FrDpJDgl3NfL/5paaoX/2kfq1uRb/138zGAhgL6COcT9SvflK/VreGXKEvhbsf\nwk7BayIiUgYNGdBnI7VJ/M7BHs6nI7WPsIiIlEHeUy5mtpnkxQCeQWovhPvMbH7BMhMRkZyUdC8X\nzclVDi1v85P61VtzzeyAbAfpTlEREU9oQBcR8YQGdBERT2hAFxHxhAZ0ERFPaEAXEfGEBnQREU9o\nQBcR8YQGdBERT2hAFxHxRNG3z5XvXXfddWE8fPhwp44s2B3bkkbLli2dcuvWrcN46dLMG4XW1dWF\n8bXXXuvUDR48OO15Lr300jCePFl711WTXXbZJYwvuugip65fv371HgcAr776ahgfeuihRcqufrpC\nFxHxhAZ0ERFPaMqlhOLTLFJ8rVq1CuPoR2EAaNOmTRj36dMn43mifXfGGWekPa5r165OecCAAWGs\nKZfKc/HFF4dx//79nboePXqEcXR6Lm7Tpk1Oedq0aQXKLne6QhcR8YQGdBERT2hAFxHxRM3OoUeX\nEALACy+8UG9cyDak9C644IIw7tmzp1M3a9asMG7WrJlT9/Of/9wpH3300WE8dOhQp27GjBlhPG/e\nPKcufl4pvPiS33PPPTeMd999d6fusMMOc8r7779/GDdqlN/17YgRI5zyjTfemNd5CkFX6CIintCA\nLiLiCa+nXHr37u2UZ86cmfbY6LK0Qt21mWmZ4vXXX1+QNiSzvffeO23dW2+9FcbvvfeeU3fNNdc4\n5dmzZ4fx6NGjnbpbb701bRtvv/12ojwlf/Fplbvvvrvoba5cubKk7SWlK3QREU9oQBcR8YQGdBER\nT9TUHHompV6qqCWNpbFgwYK0dc2bN09b9/jjjzvlKVOmpH1fdLfFuDlz5mRLUfLQrVu3MH722WdL\n3v7ll18exitWrCh5++noCl1ExBNZB3SS95FcSfK9yGvtSE4nuSj4vW1x0xQRkWxoZpkPII8AsB7A\nA2b2k+C1WwB8bmajSA4D0NbMrszaGJm5sQLL9GeL765XqCmXbH+fW5X7gRZmVrAESt2vubjqqqvC\n+KabbnLqorvkbbvttonPGd+V76GHHgrjv//9707dwIEDw/jbb79N3Ea+fO3XXXfd1SlHl5WeddZZ\nRW9/48aNTjl65/Arr7xS9PYBzDWzA7IdlPUK3cxeAvB57OWTAUwI4gkATsk5PRERKah8vxTtaGbL\ng3gFgI7pDiQ5BMCQPNuRCqV+9ZP6tbo1eJWLmVmmj2ZmNhbAWKCyPsJJw6hf/aR+rW75Duj/IVln\nZstJ1gFYmfUdFaZQc+a5LI3U7f6l9+abb+b1vvicenSnxttvv92pW7t2bRjHtwwoxby5r1q0aBHG\nI0eOdOp+9atfJTrH+vXrnfLUqVOd8qmnnproPPGlkSWaN89ZvssWJwMYFMSDADxRmHRERCRfSZYt\nPgLgNQC7k/yM5GAAowAcQ3IRgKODsoiIlFHWKRcz65+mqm+Bcymp+FRJvlMwuUy56O7Q0nvjjTfS\n1m2zzfc//ieddJJTd9555znlE044Ie15og80+OCDD3JNUQLxB5BEH7a84447Jj7Phx9+GMbRB5wA\nwPTp0/PKLX6eSqU7RUVEPKEBXUTEExrQRUQ84fVui/FlgtEnCMWfXhSdQ3/xxRfT1sXnzDM9lSgu\n2ma8jaTnif+ZNC+fv+j2C+PHj3fq2rRp45S3bNkSxrfccotTd8MNNxQ+OU/Ft7yIPm0ovqQw6bx5\n/Lb8Qw45JIzj/64aN26c6JwA8NFHH4XxN99849RFv3/57rvvnLqk238Ug67QRUQ8oQFdRMQTWXdb\nLGhjJb6VOJeHRFey6M6QBdwV0std+eKiUyeffx7fYy691atXO+WjjjoqjOMPlK4kld6vp5zi7uM3\nadKkvM4TXZp4zjnnOHWbN28O4wkTJjh1e+yxR+I2PvnkkzCeMWOGUxf9eRgxYoRTd++99yZuIweF\n2W1RRESqgwZ0ERFPaEAXEfGE18sW4/PN0SVT8eV+uSw/zLf9qExLIws1Ty75iy9NrOR580oXfaDy\nqFGF2fYpuqTxzjvvdOr23XffgrTRpUuXMI7P00dFl16Wm67QRUQ8oQFdRMQTGtBFRDzh9Tr0Qomu\nZ8+2lj16a34l35Zf6euV89WyZUunHH0izbhx4xKfp1evXk55zpw5DUusRCqhX3/605865XyfGlUt\n5s2b55QnTpwYxqNHjy5UM1qHLiJSSzSgi4h4wutli+VQydMsteA3v/mNU77tttvKlEnt+te//uWU\nzz///DCObmMRL3fs2LG4iWWxePFip7zbbrslqttrr72cuqVLlxYhu2R0hS4i4gkN6CIintCALiLi\nCS1bTCCXv6P4E1kqVSUsbyuUkSNHhvHQoUOduuiTZeJb4rZu3TqMmzRp4tT17dvXKVfL1svV1q8t\nWrQI43PPPdep22+//cJ4wYIFTl23bt3CeMiQIYnbi8/vn3TSSWG8ZMkSp65p06ZhHH8qUrQuPj6s\nXbs2cT450LJFEZFaogFdRMQTmnJJINPfUXxnxPiyrEpVbR/NM1m1alUYt2/f3qmbP39+GB977LFO\nXfQpNPEn2cTvKj3vvPManGcp+NSvmRx55JFh/Pzzzyd+X/yJSVOmTClYTkWmKRcRkVqiAV1ExBNZ\nB3SSnUnOJLmA5HySlwavtyM5neSi4Pe2xU9XRETSSXLr/2YAQ83sTZKtAMwlOR3A2QBmmNkoksMA\nDANwZfFSrUzR3RWlNE488USn3K5duzCOLxuN9s/y5cuLm5gUTaNG7rVnfIljOtGnJQHAk08+WbCc\nKlHWK3QzW25mbwbxOgDvA+gE4GQAE4LDJgA4pf4ziIhIKeS0ORfJbgD2BfAGgI5mtvWSZwWAenfW\nITkEQPKV/1IV1K9+Ur9Wt8QDOsmWACYBuMzM1kY/2pqZpVviZGZjAYwNzlGxy6CiMu2YGJ9iqdUH\nOpezX+N34m3evDmM43d8zp49uyQ5+aJS/73++te/dsoDBgxI9L6XX37ZKZdymXY5JFrlQrIJUoP5\nQ2b2WPDyf0jWBfV1AFYWJ0UREUkiySoXAhgH4H0zuzVSNRnAoCAeBOCJwqcnIiJJJZlyORTAWQDe\nJbn14Xl/BDAKwN9IDgawBMBpxUlRRESSyDqgm9ksAOluJ+6b5vWqFr2tWCrPSy+95JTXrFkTxjvs\nsINTd/PNN4fxn/70J6cuutwxLjovL+UR3VFx1KhRid8X3Zkxvkuj73SnqIiIJzSgi4h4Qrst1iOX\nv5P4MsZqeUi0T7vyXXzxxWF8++2353WOr7/+2ikfeOCBTnnhwoV5nbfUfOrXW265JYzjDy5J6qCD\nDnLKc+bMaVBOZaTdFkVEaokGdBERT2hAFxHxRE57udSK+O38vXv3DuP4E4lq9db/SnLPPfeEcfSh\nw4D7hJr4fOqWLVvCOH4rebXMmfss09LRuXPnhvGYMWOcuu7du4exli2KiEhV0oAuIuIJLVusUT4t\nb5PvqV+9pWWLIiK1RAO6iIgnNKCLiHhCA7qIiCc0oIuIeEIDuoiIJzSgi4h4QgO6iIgnNKCLiHhC\nA7qIiCdKvdviagBLAHQI4kpQi7l0LfD51K+ZqV8Lp1ZzSdS3Jd3LJWyUnJNkX4JSUC6FU0n5K5fC\nqaT8lUtmmnIREfGEBnQREU+Ua0AfW6Z266NcCqeS8lcuhVNJ+SuXDMoyh14NSE4EsNjMrktw7CwA\n95rZ+Dzayfu9kjv1q5/UrykVO+VCcn3k1xaSGyLlM8qdXzmQvDf29/Jfkl+UO69cqF/rR3I3kk+T\nXEdyNcmR5c4pF+rXHyLZjOQYkstIfkHyDpJFXVlYsQ+JNrOWW2OSHwM418yeS3c8yW3MLP1TZT1g\nZucCOHdrObgq+aZ8GeVO/fpDJLcFMB3AbQB+DcAA7FbWpHKkfq3X1QB+CqAngCYAngJwFYAbi9Vg\nxV6hZ0Pyf0k+SvIRkusAnElyIsnrIsccHfxwbS3vRPIfJFeR/IjkbxO21T64eloV/E87hWSn2GHd\nSc4h+VXQRtvI+w8l+TrJL0nOI3lEw/70AMlWAPoBmNDQc1WSGu3XwQA+NrMxZvaNmW0ws3fzPFdF\nqtF+/QWAMWb2hZmtBHAHgN/kea5EqnZAD/QD8DCA1gAezXQgyUYAngQwG0AnAMcAuIJk3wTtNALw\nFwBdkFrgvwnAmNgxA4NfOwIggNFBu50BTAYwHEA7AMMAPEayfbZGSR5JMt2NC6cCWGZmryTIv9rU\nWr8eDOATks8E0y3Pk+yZIP9qU2v9iuDc0bgbyZYokmof0GeZ2RQz22JmG7IcewiA7cxshJltNLPF\nAMYBOD1bI2a2ysz+EVw5rQUwAsCRscMmmNkCM/sawLUATidJpH5oJpvZM0Ge0wC8DeC4BO2+aGYd\n0lQPgmdX5xG11q87AegP4P8hNcBMB/AEySbZzlVlaq1fpwG4jGQHknUALgle/1G2c+WrYufQE/o0\nh2O7AuhC8svIa40BvJDtjcH/qLcBOBZAm+DlVhlyWQJgW6T+h+8KoD/JfpH6Jkh1dl5I7gzgMKR+\n+HxUa/26AcCLZvZskNfNAP4E4McA5udxvkpVa/16A1L/Sb8N4FsA9yE1n1607QKqfUCPr7n8GkDz\nSHmHSPwpgEVm1iOPdq4AsDOAXma2guQBSH0UjOocibsA+C+Az4N27zezC/NoN52BSA0ASwp4zkpS\na/36DoD9I2XDD/8OfFBT/Wpm3wC4MPgFkhcBmGNFXCte7VMucfMAnECybfAR53eRutcAbCQ5lKnl\nRI1J7kVy//pP5WiF1GqSL4K5tGvrOWYgyT1ItgBwPYC/BR33IIB+JI8J2mxGsg/JHfP5A0Y+Fo7P\n5/1Vyvd+fRDAYSSPItkYwOUAlgH4II9zVROv+zX4UreOZCOSP0Nq1ct1uZ4nF74N6OMBvI/UR6hp\nAP66tSJYInU8gF4APkbqY889ALZLcN5bkfoiZw2AVwFMreeYBwFMBLAcqY+GlwXtfozUl0HXAFgF\n4BMAQ5Hg755k79hHTiA11bI9gEkJ8vbFeHjcr2a2AKnvRO4F8EXw5zmlBpb1jYfH/QqgO4DXAaxH\narrlcjObkSD/vOlOURERT/h2hS4iUrM0oIuIeEIDuoiIJxo0oJM8juQHJBeTHFaopEREJHd5fyka\nLK/6EKlbcj9Dap1n/+Ab+3Tv0TewFcLMmP2oZNSvlUP96q3VZvY/2Q5qyBV6L6T2H/63mW1EasnR\nyQ04n4iI1C/RTYQNGdA7wb199rPgNQfJIcGuZnMa0JZUGPWrn9Sv1a3ot/6b2VgEj2rSRzh/qF/9\npH6tbg25Ql8Kdz+EnYLXRESkDBoyoM9GapP4nUk2RWpby8mFSUtERHKV95SLmW0meTGAZ5DaC+E+\nM/Npq08RkapS0r1cNCdXObS8zU/qV2/NNbMDsh2kO0VFRDyhAV1ExBMa0EVEPFHtj6ATQeohTikT\nJ0506vbZZ58w3muvvZy6LVu2FDcxkRLTFbqIiCc0oIuIeEJTLlJ1GjVyr0OGDBkSxv3790/7vjFj\nxjjlSy65pLCJiZSZrtBFRDyhAV1ExBMa0EVEPKE5dKk6HTt2dMp33XVX2mO/++67MF6zZk3RchKp\nBLpCFxHxhAZ0ERFPaMpFvLZw4cIwvu6668qXiDTIwQcf7JRHjBgRxocffrhTt379+jDetGmTU/fU\nU0855T59+oTxhAkTnLrhw4fnl2wZ6QpdRMQTGtBFRDyhAV1ExBNVP4feqlUrp3zFFVeE8YknnujU\n7bvvvmE8adIkp+6GG24I4/fff9+pi8/DFdthhx3mlI899tgwnjVrllP37LPPliSnStK6devEx06Z\nMqWImUgxtWzZMoyffvppp27mzJlhfOCBBzp18+bNK25iFUxX6CIintCALiLiiaqccolOs8yYMcOp\n23///dO+L/pAg379+jl10fKyZcucuoceeijtOcePH++Uly5dmvbYvffeO4x/8YtfpD0uvgtgs2bN\n0h7buHHjtHU+if45r7nmmrTHxafHXn755aLlJMV19dVXh/HXX3/t1J155plhvGHDhpLlVOl0hS4i\n4gkN6CIintCALiLiCZpZ6Roj82psm23cqf5XX301jDPNmcdF59C//PJLpy66RKpp06a5plgyl19+\nuVMePXp0XucxM2Y/Kpl8+zUX0acS3X333WmPGzVqlFP+4x//mOj8LVq0cMo777yzUz7nnHPCuFOn\nTonOCbjLYRcsWJD4ffmqtn7NJDo2xb9XuvPOO0udTrnNNbMDsh2kK3QREU9kHdBJ3kdyJcn3Iq+1\nIzmd5KLg97bFTVNERLJJsmxxPIA7ATwQeW0YgBlmNorksKB8ZeHTSzn99NOdctKliePGjXPqXnzx\nxTB+5JFHnLrjjjsujG+55RanrmfPnsmTLYIVK1aEcQ1+1AQA9OjRI9Fx8bt8M4lOs/35z3926s46\n66zE58nkoIMOCuN7773Xqbv55pvDePPmzQVpr5qdcMIJTvmrr74K40zTbIXSoUMHp3z//feHcXzn\nxTfffLPo+eQj6xW6mb0E4PPYyycD2LrX5AQApxQ4LxERyVG+NxZ1NLPlQbwCQMd0B5IcAmBIunqp\nTupXP6lfq1uD7xQ1M8v0bbiZjQUwFij/t+ZSOOpXP6lfq1u+A/p/SNaZ2XKSdQBWFjKpuNNOOy3x\nsR9++GEYX3DBBYnfN23atDCOzsMDQOfOncM4PrfapUuXtOeML4WLz9Gls27dOqfcvXv3MC71zo/l\nEt/SoK6uLu2xGzduDOPVq1cnbuOmm24K41zmzN99912nHL0tPbqjJwB07do1jG+88UanLroT5Dvv\nvJO4fV81adLEKUeXLRbrO4boz9lf/vIXpy46px//Pq5q59DTmAxgUBAPAvBEYdIREZF8JVm2+AiA\n1wDsTvIzkoMBjAJwDMlFAI4OyiIiUkZZp1zMrH+aqr4FziWt66+/3in/7Gc/C+P4x7T4Lor5yPTQ\niPhHr0zuuOMOp3zRRRclel/0rkQA+OabbxK36YvmzZs75UzTbsuXLw/jqVOnpj3u6KOPdsoDBgxI\ne2z8TuLf//73YfzYY485ddEpsujPJgDcdtttYXzAAe6NfmeffXYY/+EPf0ibixTP0KFDwzj+sOn4\n1Gs10J2iIiKe0IAuIuIJDegiIp6oiicWzZ071yn37fv99H18V7zossVy2GOPPcI43y0DVq1aVah0\nasJHH32Utm6XXXYJ4/h2D+3btw/j+Jx5fBnjU089lSiX6E6gADB27Ngwjs+hn3rqqWEcfToPUJtP\n4Yne6l8sffr0ccrnn39+GMe3FKnGh03rCl1ExBMa0EVEPFEVUy5xb7/9dr1xObRu3dop/+53vwvj\n/fbbL+371qxZ45Sjy9b++c9/Fii72pDpId7R5Y/RKZa4+J2aSadYGiL6oIxaedh3JjNnznTK0TuA\nr7rqKqdu5MiRic7Zq1cvpxzftTFaXrJkSaJzVjJdoYuIeEIDuoiIJzSgi4h4oirn0CtJfJ785JNP\nDuNWrVqlfV98vu7xxx8P4+jcYa3afffdy52ClNngwYPD+OGHH3bqojsxLly40Kk7/vjjwzi+ZUR0\nuwXA3fEyvhtq9OH0lbq7Ypyu0EVEPKEBXUTEExrQRUQ8oTn0HMW3573vvvuc8nbbbZf2vc8991wY\nX3rppU7d+vXrC5CdPz7/PP5c8uq05557pq178MEHw7gWt0jO5sknnwzj+JOgLrnkkjCOf98S3U65\nR48eaevi4m1Enzj2ySefJMi4/HSFLiLiCQ3oIiKeYHT5T9Ebq9KniO+zzz5hHL89OdMUS1z0iTnx\n85SambFQ5ypGvzZr1swpR5eNRXe0BIBFixaFcXTJGgD86Ec/CuNMD2J+6aWXnHLv3r0T5xo1fPhw\npxzdRTG6DA4AfvzjH4fx4sWL82ovrtL7tZKNHj3aKUd3XzziiCNKnU7cXDM7INtBukIXEfGEBnQR\nEU9oQBcR8YSWLdbjJz/5iVOOLi/LZc78jDPOcMqvvfZawxKrId9++61Tvvnmm8P4/vvvd+q6d+8e\nxq+//rpTN2PGjETt1dXVZSxnWu4WnTePP3koOm8efcI8kPlJS1J+q1evLncKOdMVuoiIJzSgi4h4\nQlMu9Yg/zDfT3X5xy5YtC+NXXnnFqYtPI0hy0Wmv+C6Wd9xxRxjHn0oU3W2PdFf0RZfsRqdtAGDy\n5MlOefz48WEcvzPxwgsvDOP4k4eiT1O6/fbbnbrvvvsOUrk05SIiImWTdUAn2ZnkTJILSM4neWnw\nejuS00kuCn5vW/x0RUQknSRX6JsBDDWzPQEcDOC3JPcEMAzADDPrDmBGUBYRkTLJOoduZssBLA/i\ndSTfB9AJwMkAegeHTQDwAoAri5JlCZx66qlhPGbMmMTvi86ZA+5ujJ9++mnDExMAwJYtW8L4rrvu\ncurWrVsXxlde6f4IRrcJyGWbi+ht3/WVk9qwYUMYa868ssWXqk6dOrVMmeQvpy9FSXYDsC+ANwB0\nDAZ7AFgBoGOa9wwBMCT/FKUSqV/9pH6tbokHdJItAUwCcJmZrY2uGDAzS7eRj5mNBTA2OEdNbfbj\nM/Wrn9Sv1S3RgE6yCVKD+UNm9ljw8n9I1pnZcpJ1AFYWK8lS6N+/fxi3bNky8fseeOABpzxnzpyC\n5ST1i06/AMCECRPC+NFHH3XqdthhhzB+5plnnLr4UsVCmD9/vlOOtymV66CDDnLK06ZNK1Mm+Uuy\nyoUAxgF438xujVRNBjAoiAcBeKLw6YmISFJJrtAPBXAWgHdJzgte+yOAUQD+RnIwgCUATkvzfhER\nKYEkq1xmAUi3aX7fwqYjIiL5qtlb/+O3bx9zzDF5nSd+i7iUV3x7hY8//jiMe/bs6dQNGDAgjI86\n6iinbuDAgU45+jDwXXfd1al7+eWXwzi6DQEArFq1KkHWUi5NmzYN4/gTpUr5NLdC0a3/IiKe0IAu\nIuKJmppyadOmTRjfc889Tl3z5s0TneOXv/ylU37rrbcanpiUxObNm51ydMlpfPnp2WefXYqUpMza\ntWsXxi1atChjJoWhK3QREU9oQBcR8YQGdBERT9TUHPr2228fxocffnji940ePTqMn376aadu06ZN\nDU9MRMpixYoVYfzVV185dfGlzdVAV+giIp7QgC4i4omamnJJavbs2U45+tAEPaRAxE/RXTsBYO+9\n9y5TJvnTFbqIiCc0oIuIeEIDuoiIJ1jKHcXK/UirDh06hPHMmTOduugtwH37ursCL1y4sLiJlYGZ\npdsSOWfsusj6AAACh0lEQVTl7lf5nvrVW3PN7IBsB+kKXUTEExrQRUQ8UVNTLvI9fTT3k/rVW5py\nERGpJRrQRUQ8oQFdRMQTpb71fzWAJQA6BHElqMVcuhb4fOrXzNSvhVOruSTq25J+KRo2Ss5JMsFf\nCsqlcCopf+VSOJWUv3LJTFMuIiKe0IAuIuKJcg3oY8vUbn2US+FUUv7KpXAqKX/lkkFZ5tBFRKTw\nNOUiIuIJDegiIp4o6YBO8jiSH5BcTHJYKdsO2r+P5EqS70Vea0dyOslFwe9tS5BHZ5IzSS4gOZ/k\npeXKpRDUr04u3vSt+tXJpSr6tWQDOsnGAP4M4P8A2BNAf5J7lqr9wHgAx8VeGwZghpl1BzAjKBfb\nZgBDzWxPAAcD+G3wd1GOXBpE/foDXvSt+vUHqqNfzawkvwAcAuCZSPkqAFeVqv1Iu90AvBcpfwCg\nLojrAHxQhpyeAHBMJeSiflXfql+rt19LOeXSCcCnkfJnwWvl1tHMlgfxCgAdS9k4yW4A9gXwRrlz\nyZP6NY0q71v1axqV3K/6UjTCUv/NlmwdJ8mWACYBuMzM1pYzF5+V4+9SfVt86tcfKuWAvhRA50h5\np+C1cvsPyToACH5fWYpGSTZB6gfjITN7rJy5NJD6NcaTvlW/xlRDv5ZyQJ8NoDvJnUk2BXA6gMkl\nbD+dyQAGBfEgpObGiookAYwD8L6Z3VrOXApA/RrhUd+qXyOqpl9L/EXC8QA+BPAvAFeX4YuMRwAs\nB7AJqTnBwQDaI/Xt9CIAzwFoV4I8DkPqo9k7AOYFv44vRy7qV/Wt+tWfftWt/yIintCXoiIintCA\nLiLiCQ3oIiKe0IAuIuIJDegiIp7QgC4i4gkN6CIinvj/Y06YonEBGQsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11185acc0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plotImages(X_valid, Y_pred, shuffle=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can also measure the the model accuracy on the validation set i.e. the percentage of the correctly labelled validation samples."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KNN accuracy = 96.3%\n"
     ]
    }
   ],
   "source": [
    "print(\"KNN accuracy = {}%\".format(np.mean(Y_valid == Y_pred)*100))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's try another k value."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "del neigh # delete the previous KNN to free up memory\n",
    "neigh = KNeighborsClassifier(n_neighbors=3, n_jobs=-1) # n_jobs sets the number of parallel jobs. -1 means max possible.\n",
    "# KNN expects a feature vector for each sample. We reshape the training data back to the flattened form\n",
    "neigh.fit(X_train.reshape(-1, 784), Y_train)\n",
    "Y_pred = neigh.predict(X_valid.reshape(-1, 784))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KNN accuracy = 96.75%\n"
     ]
    }
   ],
   "source": [
    "print(\"KNN accuracy = {}%\".format(np.mean(Y_valid == Y_pred)*100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "del neigh # delete the KNN model to free up memory"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here is the workflow again:\n",
    "![alt text](figs/workflow.png)\n",
    "Did we missed something?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# How to do \"multilabel\" classification?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Excercise**: Apply multiclass SVM from sklearn to this dataset.\n",
    "\n",
    "NOTES:\n",
    "- You should apply PCA first\n",
    "- You can use `LinearSVC` class.\n",
    "- Google is your friend!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(40000, 10)\n",
      "SVM accuracy = 62.5%\n"
     ]
    }
   ],
   "source": [
    "# Let's use 10 features ...\n",
    "\n",
    "pca = PCA(n_components=10)\n",
    "pca.fit(X_train.reshape(-1, 784))\n",
    "X_PCA = pca.transform(X_train.reshape(-1, 784))\n",
    "X_PCA_valid = pca.transform(X_valid.reshape(-1,784))\n",
    "print(X_PCA.shape)\n",
    "clf = LinearSVC(random_state=0)\n",
    "\n",
    "clf.fit(X_PCA, Y_train)\n",
    "Y_pred = clf.predict(X_PCA_valid)\n",
    "print(\"SVM accuracy = {}%\".format(np.mean(Y_valid == Y_pred)*100))"
   ]
  }
 ],
 "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": "3.6.3"
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}