zerosteer.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Import the required libraries and initial view"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import os\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df = pd.read_csv('/Users/sujith/trainData/driving_log.csv', header=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.columns=[\"centre\", \"left\", \"right\", \"steer\", \"x1\", \"x2\", \"x3\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>centre</th>\n",
       "      <th>left</th>\n",
       "      <th>right</th>\n",
       "      <th>steer</th>\n",
       "      <th>x1</th>\n",
       "      <th>x2</th>\n",
       "      <th>x3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>/Users/sujith/trainData/IMG/center_2019_02_15_...</td>\n",
       "      <td>/Users/sujith/trainData/IMG/left_2019_02_15_12...</td>\n",
       "      <td>/Users/sujith/trainData/IMG/right_2019_02_15_1...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>16.22692</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>/Users/sujith/trainData/IMG/center_2019_02_15_...</td>\n",
       "      <td>/Users/sujith/trainData/IMG/left_2019_02_15_12...</td>\n",
       "      <td>/Users/sujith/trainData/IMG/right_2019_02_15_1...</td>\n",
       "      <td>-0.15</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>17.51071</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>/Users/sujith/trainData/IMG/center_2019_02_15_...</td>\n",
       "      <td>/Users/sujith/trainData/IMG/left_2019_02_15_12...</td>\n",
       "      <td>/Users/sujith/trainData/IMG/right_2019_02_15_1...</td>\n",
       "      <td>-0.40</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>18.65486</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>/Users/sujith/trainData/IMG/center_2019_02_15_...</td>\n",
       "      <td>/Users/sujith/trainData/IMG/left_2019_02_15_12...</td>\n",
       "      <td>/Users/sujith/trainData/IMG/right_2019_02_15_1...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>19.84270</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>/Users/sujith/trainData/IMG/center_2019_02_15_...</td>\n",
       "      <td>/Users/sujith/trainData/IMG/left_2019_02_15_12...</td>\n",
       "      <td>/Users/sujith/trainData/IMG/right_2019_02_15_1...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>21.03341</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                              centre  \\\n",
       "0  /Users/sujith/trainData/IMG/center_2019_02_15_...   \n",
       "1  /Users/sujith/trainData/IMG/center_2019_02_15_...   \n",
       "2  /Users/sujith/trainData/IMG/center_2019_02_15_...   \n",
       "3  /Users/sujith/trainData/IMG/center_2019_02_15_...   \n",
       "4  /Users/sujith/trainData/IMG/center_2019_02_15_...   \n",
       "\n",
       "                                                left  \\\n",
       "0  /Users/sujith/trainData/IMG/left_2019_02_15_12...   \n",
       "1  /Users/sujith/trainData/IMG/left_2019_02_15_12...   \n",
       "2  /Users/sujith/trainData/IMG/left_2019_02_15_12...   \n",
       "3  /Users/sujith/trainData/IMG/left_2019_02_15_12...   \n",
       "4  /Users/sujith/trainData/IMG/left_2019_02_15_12...   \n",
       "\n",
       "                                               right  steer   x1  x2        x3  \n",
       "0  /Users/sujith/trainData/IMG/right_2019_02_15_1...   0.00  1.0   0  16.22692  \n",
       "1  /Users/sujith/trainData/IMG/right_2019_02_15_1...  -0.15  1.0   0  17.51071  \n",
       "2  /Users/sujith/trainData/IMG/right_2019_02_15_1...  -0.40  1.0   0  18.65486  \n",
       "3  /Users/sujith/trainData/IMG/right_2019_02_15_1...   0.00  1.0   0  19.84270  \n",
       "4  /Users/sujith/trainData/IMG/right_2019_02_15_1...   0.00  1.0   0  21.03341  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3466, 7)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAD8CAYAAAB+UHOxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10a618cc0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.steer.hist()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Verify that IMG folder and the CSV file are in sync"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Received DF with rows =  3466\n",
      "Done...\n"
     ]
    }
   ],
   "source": [
    "newdf = pd.read_csv('/Users/sujith/trainData/driving_log.csv', header=None)\n",
    "orig_count = len(newdf)\n",
    "print(\"Received DF with rows = \", orig_count)\n",
    "\n",
    "for i in range(orig_count):\n",
    "    for j in range(3):\n",
    "        path = newdf.iloc[i][j]\n",
    "        if not os.path.exists(path):\n",
    "            print(\" *************   Not found: \", path)\n",
    "\n",
    "print (\"Done...\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Dropping the near-zero predictions\n",
    "\n",
    "This function takes in the range of steering angle (either side of 0-degree) and the proportion of the images to be deleted as parms. It identifies all the images with steering angle less than the provided range, calculate the count of images to be deleted (based on parm provided) and randomly deletes the portion of images.\n",
    "\n",
    "**Note:** Apart from removing the entries from the CSV file, I am also deleting the images from the IMG folder so that I could save on both time and bandwidth when uploading the images to Udacity platform for training."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def drop_zero_steering(dataframe, drop_zero_prob=0.75, drop_range=0.05):\n",
    "    orig_count = len(dataframe)\n",
    "    print(\"Received DF with rows = \", orig_count)\n",
    "    near_zero_idx = np.where(abs(pd.to_numeric(dataframe['steer'])) < drop_range)[0]\n",
    "    print(\"Count of near_zero_idx: \", len(near_zero_idx))\n",
    "    delete_count = int(drop_zero_prob * len(near_zero_idx))\n",
    "    print(\"Count identified for deletion: \", delete_count)\n",
    "    delete_indices = np.random.choice(near_zero_idx, delete_count, replace=False)\n",
    "    print(\"Count of identified indices for deletion: \", len(delete_indices))\n",
    "    for i in range(delete_count):\n",
    "        for j in range(3):\n",
    "            path = dataframe.iloc[delete_indices[i]][j]\n",
    "            if os.path.exists(path):\n",
    "                os.remove(path)\n",
    "                # print(path)\n",
    "            else:\n",
    "                print(\" *************   Not found: \", path)\n",
    "\n",
    "    # Update steering angles array.\n",
    "    dataframe.drop(dataframe.index[delete_indices], inplace=True)\n",
    "\n",
    "    new_count = len(dataframe)\n",
    "    print(\"{} rows dropped.\".format(orig_count - new_count))\n",
    "    print(\"Output DF has rows = \", new_count)\n",
    "    return dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Received DF with rows =  3466\n",
      "Count of near_zero_idx:  2721\n",
      "Count identified for deletion:  2040\n",
      "Count of identified indices for deletion:  2040\n",
      "2040 rows dropped.\n",
      "Output DF has rows =  1426\n"
     ]
    }
   ],
   "source": [
    "temp = drop_zero_steering(df)\n",
    "temp.to_csv(\"/Users/sujith/trainData/driving_log_norm.csv\", index=None, header=False)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Verify that IMG folder and the CSV file are in sync (after deletion)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Received DF with rows =  1426\n",
      "Done...\n"
     ]
    }
   ],
   "source": [
    "newdf = pd.read_csv('/Users/sujith/trainData/driving_log_norm.csv', header=None)\n",
    "orig_count = len(newdf)\n",
    "print(\"Received DF with rows = \", orig_count)\n",
    "\n",
    "for i in range(orig_count):\n",
    "    for j in range(3):\n",
    "        path = newdf.iloc[i][j]\n",
    "        if not os.path.exists(path):\n",
    "            print(\" *************   Not found: \", path)\n",
    "\n",
    "print (\"Done...\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Visualize the effect of removal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAD8CAYAAAB5Pm/hAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x106179c18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "temp.steer.hist()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ignore below steps.\n",
    "There are two steps which are actually performed in the model pipeline. They are:\n",
    "1. Adjusting the steering angle for left and right camera images\n",
    "2. Data augmentation by horizontal flip of images. \n",
    "\n",
    "As these are part of pipeline, it is not performed locally, but I am running the below CELLs just to capture the effect of these steps in a histogram."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def correct_steer(temp):\n",
    "    measurements = []\n",
    "    correction = 0.2\n",
    "\n",
    "    for row in range(len(temp)):\n",
    "        for i in range(3):\n",
    "            if i == 0:\n",
    "                measurements.append(float(temp.iloc[row,3]))\n",
    "            elif i == 1: # left camera\n",
    "                measurements.append(float(temp.iloc[row,3]) + correction)\n",
    "            else:        # right camera\n",
    "                measurements.append(float(temp.iloc[row,3]) - correction)\n",
    "    return measurements\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYEAAAD8CAYAAACRkhiPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nHUEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10a6ef2e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "m = correct_steer(temp)\n",
    "plt.hist(m)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def aug_imgs(measurements):\n",
    "    aug_measus = []\n",
    "    for measu in measurements:\n",
    "        aug_measus.append(measu)\n",
    "        aug_measus.append(-1.0 * measu)\n",
    "    return aug_measus\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAD8CAYAAAB+UHOxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAEoRJREFUeJzt3W2MneV95/Hvr4aQalsVEwbWtZ01\nbb1qyUo1aNZBmzdZyIKhUky6RYIXxRuxciOB1ErVqk77gjZZtKm0LRJSikSKN2bVhdC0FW7ilnVJ\noigveBh2HYNxqSeEhqktPK0JbZRddmH/+2Iu7x5gHs48nRP7+n6kW+e+//d13+e65szMb+6HcyZV\nhSSpPz807g5IksbDAJCkThkAktQpA0CSOmUASFKnDABJ6pQBIEmdMgAkqVMGgCR16oJxd2Axl156\naW3btm3c3ZCkc8qzzz77t1U1sVS7H+gA2LZtG1NTU+PuhiSdU5L89TDtPAUkSZ0yACSpUwaAJHXK\nAJCkThkAktQpA0CSOmUASFKnDABJ6tSSAZDkvUmeTvLNJMeS/Farfz7Jt5McadOOVk+S+5JMJzma\n5OqBfe1JcqJNe9ZvWJKkpQzzTuA3gGur6ntJLgS+keTP2rp/V1VffEf7G4HtbfogcD/wwSSXAHcD\nk0ABzyY5WFWvrcVApFHbtu/LY3nelz/zc2N5Xp1/ljwCqDnfa4sXtqkW2WQ38FDb7kng4iSbgBuA\nw1V1pv3SPwzsWl33JUkrNdQ1gCQbkhwBTjP3S/yptuqedprn3iQXtdpm4JWBzWdabaG6JGkMhgqA\nqnqrqnYAW4CdSf4Z8Engp4F/DlwC/Fprnvl2sUj9bZLsTTKVZGp2dnaY7kmSVmBZdwFV1XeBrwG7\nqupUO83zBvCfgJ2t2QywdWCzLcDJRervfI4HqmqyqiYnJpb8NFNJ0goNcxfQRJKL2/wPAx8B/rKd\n1ydJgJuB59smB4Hb291A1wCvV9Up4HHg+iQbk2wErm81SdIYDHMX0CbgQJINzAXGo1X1pSRfSTLB\n3KmdI8AnWvtDwE3ANPB94OMAVXUmyaeBZ1q7T1XVmbUbiiRpOZYMgKo6Clw1T/3aBdoXcOcC6/YD\n+5fZR0nSOvCdwJLUKQNAkjplAEhSpwwASeqUASBJnTIAJKlTBoAkdcoAkKROGQCS1CkDQJI6ZQBI\nUqcMAEnqlAEgSZ0yACSpUwaAJHXKAJCkThkAktQpA0CSOmUASFKnDABJ6tSSAZDkvUmeTvLNJMeS\n/FarX5HkqSQnknwhyXta/aK2PN3WbxvY1ydb/cUkN6zXoCRJSxvmCOAN4Nqq+llgB7AryTXAbwP3\nVtV24DXgjtb+DuC1qvop4N7WjiRXArcCHwB2Ab+XZMNaDkaSNLwlA6DmfK8tXtimAq4FvtjqB4Cb\n2/zutkxbf12StPojVfVGVX0bmAZ2rskoJEnLNtQ1gCQbkhwBTgOHgW8B362qN1uTGWBzm98MvALQ\n1r8OvG+wPs82kqQRGyoAquqtqtoBbGHur/afma9Ze8wC6xaqv02SvUmmkkzNzs4O0z1J0gos6y6g\nqvou8DXgGuDiJBe0VVuAk21+BtgK0Nb/GHBmsD7PNoPP8UBVTVbV5MTExHK6J0lahmHuAppIcnGb\n/2HgI8Bx4KvAL7Rme4DH2vzBtkxb/5Wqqla/td0ldAWwHXh6rQYiSVqeC5ZuwibgQLtj54eAR6vq\nS0leAB5J8u+B/w482No/CPznJNPM/eV/K0BVHUvyKPAC8CZwZ1W9tbbDkSQNa8kAqKqjwFXz1F9i\nnrt4qup/ArcssK97gHuW301J0lrzncCS1CkDQJI6ZQBIUqcMAEnqlAEgSZ0yACSpUwaAJHXKAJCk\nThkAktQpA0CSOmUASFKnDABJ6pQBIEmdMgAkqVMGgCR1ygCQpE4ZAJLUKQNAkjplAEhSpwwASeqU\nASBJnVoyAJJsTfLVJMeTHEvyy63+m0n+JsmRNt00sM0nk0wneTHJDQP1Xa02nWTf+gxJkjSMC4Zo\n8ybwq1X135L8KPBsksNt3b1V9R8HGye5ErgV+ADw48BfJPmnbfVngX8FzADPJDlYVS+sxUAkScuz\nZABU1SngVJv/hyTHgc2LbLIbeKSq3gC+nWQa2NnWTVfVSwBJHmltDQBJGoNlXQNIsg24Cniqle5K\ncjTJ/iQbW20z8MrAZjOttlD9nc+xN8lUkqnZ2dnldE+StAxDB0CSHwH+CPiVqvp74H7gJ4EdzB0h\n/M7ZpvNsXovU316oeqCqJqtqcmJiYtjuSZKWaZhrACS5kLlf/n9QVX8MUFWvDqz/HPCltjgDbB3Y\nfAtwss0vVJckjdgwdwEFeBA4XlW/O1DfNNDsY8Dzbf4gcGuSi5JcAWwHngaeAbYnuSLJe5i7UHxw\nbYYhSVquYY4APgT8IvBckiOt9uvAbUl2MHca52XglwCq6liSR5m7uPsmcGdVvQWQ5C7gcWADsL+q\njq3hWCRJyzDMXUDfYP7z94cW2eYe4J556ocW206SNDq+E1iSOmUASFKnDABJ6pQBIEmdMgAkqVMG\ngCR1ygCQpE4ZAJLUKQNAkjplAEhSpwwASeqUASBJnTIAJKlTBoAkdcoAkKROGQCS1CkDQJI6ZQBI\nUqcMAEnqlAEgSZ1aMgCSbE3y1STHkxxL8sutfkmSw0lOtMeNrZ4k9yWZTnI0ydUD+9rT2p9Ismf9\nhiVJWsowRwBvAr9aVT8DXAPcmeRKYB/wRFVtB55oywA3AtvbtBe4H+YCA7gb+CCwE7j7bGhIkkbv\ngqUaVNUp4FSb/4ckx4HNwG7gw63ZAeBrwK+1+kNVVcCTSS5Osqm1PVxVZwCSHAZ2AQ+v4XjUmW37\nvjzuLkjnrGVdA0iyDbgKeAq4vIXD2ZC4rDXbDLwysNlMqy1Uf+dz7E0ylWRqdnZ2Od2TJC3D0AGQ\n5EeAPwJ+par+frGm89RqkfrbC1UPVNVkVU1OTEwM2z1J0jINFQBJLmTul/8fVNUft/Kr7dQO7fF0\nq88AWwc23wKcXKQuSRqDYe4CCvAgcLyqfndg1UHg7J08e4DHBuq3t7uBrgFeb6eIHgeuT7KxXfy9\nvtUkSWOw5EVg4EPALwLPJTnSar8OfAZ4NMkdwHeAW9q6Q8BNwDTwfeDjAFV1JsmngWdau0+dvSAs\nSRq9Ye4C+gbzn78HuG6e9gXcucC+9gP7l9NBSdL68J3AktQpA0CSOmUASFKnDABJ6pQBIEmdMgAk\nqVMGgCR1ygCQpE4ZAJLUKQNAkjplAEhSpwwASeqUASBJnTIAJKlTBoAkdcoAkKROGQCS1CkDQJI6\nZQBIUqcMAEnq1JIBkGR/ktNJnh+o/WaSv0lypE03Daz7ZJLpJC8muWGgvqvVppPsW/uhSJKWY5gj\ngM8Du+ap31tVO9p0CCDJlcCtwAfaNr+XZEOSDcBngRuBK4HbWltJ0phcsFSDqvp6km1D7m838EhV\nvQF8O8k0sLOtm66qlwCSPNLavrDsHkuS1sRqrgHcleRoO0W0sdU2A68MtJlptYXq75Jkb5KpJFOz\ns7Or6J4kaTErDYD7gZ8EdgCngN9p9czTthapv7tY9UBVTVbV5MTExAq7J0laypKngOZTVa+enU/y\nOeBLbXEG2DrQdAtwss0vVJckjcGKjgCSbBpY/Bhw9g6hg8CtSS5KcgWwHXgaeAbYnuSKJO9h7kLx\nwZV3W5K0WkseASR5GPgwcGmSGeBu4MNJdjB3Gudl4JcAqupYkkeZu7j7JnBnVb3V9nMX8DiwAdhf\nVcfWfDSSpKENcxfQbfOUH1yk/T3APfPUDwGHltU7SdK68Z3AktQpA0CSOmUASFKnDABJ6pQBIEmd\nMgAkqVMGgCR1ygCQpE4ZAJLUKQNAkjplAEhSpwwASeqUASBJnTIAJKlTBoAkdcoAkKROGQCS1CkD\nQJI6ZQBIUqcMAEnq1JIBkGR/ktNJnh+oXZLkcJIT7XFjqyfJfUmmkxxNcvXANnta+xNJ9qzPcCRJ\nwxrmCODzwK531PYBT1TVduCJtgxwI7C9TXuB+2EuMIC7gQ8CO4G7z4aGJGk8lgyAqvo6cOYd5d3A\ngTZ/ALh5oP5QzXkSuDjJJuAG4HBVnamq14DDvDtUJEkjtNJrAJdX1SmA9nhZq28GXhloN9NqC9Ul\nSWOy1heBM0+tFqm/ewfJ3iRTSaZmZ2fXtHOSpP9vpQHwaju1Q3s83eozwNaBdluAk4vU36WqHqiq\nyaqanJiYWGH3JElLWWkAHATO3smzB3hsoH57uxvoGuD1doroceD6JBvbxd/rW02SNCYXLNUgycPA\nh4FLk8wwdzfPZ4BHk9wBfAe4pTU/BNwETAPfBz4OUFVnknwaeKa1+1RVvfPCsiRphJYMgKq6bYFV\n183TtoA7F9jPfmD/snonSVo3vhNYkjplAEhSpwwASeqUASBJnTIAJKlTBoAkdcoAkKROGQCS1Kkl\n3wgmDWPbvi+PuwvdGOfX+uXP/NzYnltrzyMASeqUASBJnTIAJKlTBoAkdcoAkKROGQCS1CkDQJI6\nZQBIUqcMAEnqlAEgSZ0yACSpU6sKgCQvJ3kuyZEkU612SZLDSU60x42tniT3JZlOcjTJ1WsxAEnS\nyqzFEcC/rKodVTXZlvcBT1TVduCJtgxwI7C9TXuB+9fguSVJK7Qep4B2Awfa/AHg5oH6QzXnSeDi\nJJvW4fklSUNYbQAU8F+TPJtkb6tdXlWnANrjZa2+GXhlYNuZVpMkjcFq/x/Ah6rqZJLLgMNJ/nKR\ntpmnVu9qNBckewHe//73r7J7kqSFrOoIoKpOtsfTwJ8AO4FXz57aaY+nW/MZYOvA5luAk/Ps84Gq\nmqyqyYmJidV0T5K0iBUHQJJ/lORHz84D1wPPAweBPa3ZHuCxNn8QuL3dDXQN8PrZU0WSpNFbzSmg\ny4E/SXJ2P/+lqv48yTPAo0nuAL4D3NLaHwJuAqaB7wMfX8VzawH+a0atp3F9f/mvKNfHigOgql4C\nfnae+t8B181TL+DOlT6fJGlt+U5gSeqUASBJnTIAJKlTBoAkdcoAkKROGQCS1CkDQJI6ZQBIUqcM\nAEnqlAEgSZ0yACSpUwaAJHXKAJCkTq32P4JpHn4ks7S2xvkzdT5/FLVHAJLUKQNAkjplAEhSpwwA\nSerUeX0R2IuxklbrfP4/yB4BSFKnRh4ASXYleTHJdJJ9o35+SdKckQZAkg3AZ4EbgSuB25JcOco+\nSJLmjPoIYCcwXVUvVdX/Ah4Bdo+4D5IkRh8Am4FXBpZnWk2SNGKjvgso89TqbQ2SvcDetvi9JC+u\ne69W51Lgb8fdiTHpeezQ9/h7HjuMYPz57VVt/k+GaTTqAJgBtg4sbwFODjaoqgeAB0bZqdVIMlVV\nk+Puxzj0PHboe/w9jx3On/GP+hTQM8D2JFckeQ9wK3BwxH2QJDHiI4CqejPJXcDjwAZgf1UdG2Uf\nJElzRv5O4Ko6BBwa9fOuo3PmdNU66Hns0Pf4ex47nCfjT1Ut3UqSdN7xoyAkqVMGwDIluSXJsST/\nJ8mCdwGcjx95keSSJIeTnGiPGxdo91aSI2065y/yL/VaJrkoyRfa+qeSbBt9L9fHEGP/N0lmB17v\nfzuOfq6HJPuTnE7y/ALrk+S+9rU5muTqUfdxtQyA5Xse+Hng6ws1OI8/8mIf8ERVbQeeaMvz+R9V\ntaNNHx1d99bekK/lHcBrVfVTwL3A6u7g/gGxjO/jLwy83r8/0k6ur88DuxZZfyOwvU17gftH0Kc1\nZQAsU1Udr6ql3px2vn7kxW7gQJs/ANw8xr6MyjCv5eDX5YvAdUnme9PjueZ8/T4eSlV9HTizSJPd\nwEM150ng4iSbRtO7tWEArI/z9SMvLq+qUwDt8bIF2r03yVSSJ5Oc6yExzGv5/9pU1ZvA68D7RtK7\n9TXs9/G/bqdAvphk6zzrz1fn/M/5ef0PYVYqyV8A/3ieVb9RVY8Ns4t5aufE7VaLjX0Zu3l/VZ1M\n8hPAV5I8V1XfWpsejtwwr+U5+3ovYZhx/SnwcFW9keQTzB0JXbvuPfvBcM6/7gbAPKrqI6vcxZIf\nefGDarGxJ3k1yaaqOtUOdU8vsI+T7fGlJF8DrgLO1QAY5rU822YmyQXAj7H4qYNzxTAf3fJ3A4uf\n4zy5/jGkc/bn/CxPAa2P8/UjLw4Ce9r8HuBdR0NJNia5qM1fCnwIeGFkPVx7w7yWg1+XXwC+UufH\nG2yWHPs7znl/FDg+wv6N20Hg9nY30DXA62dPkZ4zqsppGRPwMeaS/w3gVeDxVv9x4NBAu5uAv2Lu\nL9/fGHe/12js72Pu7p8T7fGSVp8Efr/N/wvgOeCb7fGOcfd7Dcb9rtcS+BTw0Tb/XuAPgWngaeAn\nxt3nEY79PwDH2uv9VeCnx93nNRz7w8Ap4H+3n/k7gE8An2jrw9xdUt9q3+uT4+7zciffCSxJnfIU\nkCR1ygCQpE4ZAJLUKQNAkjplAEhSpwwASeqUASBJnTIAJKlT/xe4aAtPdmvA3gAAAABJRU5ErkJg\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1129311d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "a = aug_imgs(m)\n",
    "plt.hist(a)\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}