EDA.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 27,
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "from fastai.data.external import untar_data, URLs\n",
    "from os.path import join\n",
    "from PIL import Image"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "outputs": [],
   "source": [
    "datapath = untar_data(URLs.IMAGEWOOF)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "outputs": [
    {
     "data": {
      "text/plain": "                                   path noisy_labels_0 noisy_labels_1  \\\n0   train/n02115641/n02115641_3995.JPEG      n02115641      n02115641   \n1    train/n02115641/n02115641_843.JPEG      n02115641      n02115641   \n2   train/n02115641/n02115641_2953.JPEG      n02115641      n02115641   \n3   train/n02115641/n02115641_6458.JPEG      n02115641      n02115641   \n4  train/n02115641/n02115641_19414.JPEG      n02115641      n02115641   \n\n  noisy_labels_5 noisy_labels_25 noisy_labels_50  is_valid  \n0      n02115641       n02115641       n02115641     False  \n1      n02105641       n02115641       n02088364     False  \n2      n02115641       n02111889       n02099601     False  \n3      n02115641       n02093754       n02115641     False  \n4      n02115641       n02115641       n02088364     False  ",
      "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>path</th>\n      <th>noisy_labels_0</th>\n      <th>noisy_labels_1</th>\n      <th>noisy_labels_5</th>\n      <th>noisy_labels_25</th>\n      <th>noisy_labels_50</th>\n      <th>is_valid</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>train/n02115641/n02115641_3995.JPEG</td>\n      <td>n02115641</td>\n      <td>n02115641</td>\n      <td>n02115641</td>\n      <td>n02115641</td>\n      <td>n02115641</td>\n      <td>False</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>train/n02115641/n02115641_843.JPEG</td>\n      <td>n02115641</td>\n      <td>n02115641</td>\n      <td>n02105641</td>\n      <td>n02115641</td>\n      <td>n02088364</td>\n      <td>False</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>train/n02115641/n02115641_2953.JPEG</td>\n      <td>n02115641</td>\n      <td>n02115641</td>\n      <td>n02115641</td>\n      <td>n02111889</td>\n      <td>n02099601</td>\n      <td>False</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>train/n02115641/n02115641_6458.JPEG</td>\n      <td>n02115641</td>\n      <td>n02115641</td>\n      <td>n02115641</td>\n      <td>n02093754</td>\n      <td>n02115641</td>\n      <td>False</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>train/n02115641/n02115641_19414.JPEG</td>\n      <td>n02115641</td>\n      <td>n02115641</td>\n      <td>n02115641</td>\n      <td>n02115641</td>\n      <td>n02088364</td>\n      <td>False</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv(join(datapath, 'noisy_imagewoof.csv'))\n",
    "data.head()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "outputs": [
    {
     "data": {
      "text/plain": "path               0\nnoisy_labels_0     0\nnoisy_labels_1     0\nnoisy_labels_5     0\nnoisy_labels_25    0\nnoisy_labels_50    0\nis_valid           0\ndtype: int64"
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.isna().sum()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "outputs": [
    {
     "data": {
      "text/plain": "((9025, 7), (3929, 7))"
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data = data[data.path.apply(lambda x: x.startswith('train'))]\n",
    "val_data = data[data.path.apply(lambda x: x.startswith('val'))]\n",
    "train_data.shape, val_data.shape"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "outputs": [
    {
     "data": {
      "text/plain": "           data  train_split  val_split\nn02115641  1350          940        410\nn02086240  1350          941        409\nn02088364  1350          932        418\nn02087394  1350          942        408\nn02105641  1350          928        422\nn02111889  1350          921        429\nn02099601  1350          949        401\nn02096294  1350          943        407\nn02093754  1350          949        401\nn02089973   804          580        224",
      "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>data</th>\n      <th>train_split</th>\n      <th>val_split</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>n02115641</th>\n      <td>1350</td>\n      <td>940</td>\n      <td>410</td>\n    </tr>\n    <tr>\n      <th>n02086240</th>\n      <td>1350</td>\n      <td>941</td>\n      <td>409</td>\n    </tr>\n    <tr>\n      <th>n02088364</th>\n      <td>1350</td>\n      <td>932</td>\n      <td>418</td>\n    </tr>\n    <tr>\n      <th>n02087394</th>\n      <td>1350</td>\n      <td>942</td>\n      <td>408</td>\n    </tr>\n    <tr>\n      <th>n02105641</th>\n      <td>1350</td>\n      <td>928</td>\n      <td>422</td>\n    </tr>\n    <tr>\n      <th>n02111889</th>\n      <td>1350</td>\n      <td>921</td>\n      <td>429</td>\n    </tr>\n    <tr>\n      <th>n02099601</th>\n      <td>1350</td>\n      <td>949</td>\n      <td>401</td>\n    </tr>\n    <tr>\n      <th>n02096294</th>\n      <td>1350</td>\n      <td>943</td>\n      <td>407</td>\n    </tr>\n    <tr>\n      <th>n02093754</th>\n      <td>1350</td>\n      <td>949</td>\n      <td>401</td>\n    </tr>\n    <tr>\n      <th>n02089973</th>\n      <td>804</td>\n      <td>580</td>\n      <td>224</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "total_cnt = pd.value_counts(data.path.apply(lambda x: x.split('/')[1]))\n",
    "train_cnt = pd.value_counts(train_data.path.apply(lambda x: x.split('/')[1]))\n",
    "val_cnt = pd.value_counts(val_data.path.apply(lambda x: x.split('/')[1]))\n",
    "pd.concat([total_cnt, train_cnt, val_cnt], axis=1, keys=['data', 'train_split', 'val_split'])"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "outputs": [],
   "source": [
    "def get_width(img_path):\n",
    "    image = Image.open(img_path)\n",
    "    return image.size[0]\n",
    "def get_height(img_path):\n",
    "    image = Image.open(img_path)\n",
    "    return image.size[1]\n",
    "def get_mode(img_path):\n",
    "    image = Image.open(img_path)\n",
    "    return image.mode"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "outputs": [
    {
     "data": {
      "text/plain": "   width  height mode        cls\n0    500     375  RGB  n02115641\n1    500     375  RGB  n02115641\n2    360     284  RGB  n02115641\n3    120      90  RGB  n02115641\n4    500     375  RGB  n02115641",
      "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>width</th>\n      <th>height</th>\n      <th>mode</th>\n      <th>cls</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>500</td>\n      <td>375</td>\n      <td>RGB</td>\n      <td>n02115641</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>500</td>\n      <td>375</td>\n      <td>RGB</td>\n      <td>n02115641</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>360</td>\n      <td>284</td>\n      <td>RGB</td>\n      <td>n02115641</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>120</td>\n      <td>90</td>\n      <td>RGB</td>\n      <td>n02115641</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>500</td>\n      <td>375</td>\n      <td>RGB</td>\n      <td>n02115641</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "widths = data.path.apply(lambda x: get_width(join(datapath, x)))\n",
    "heights = data.path.apply(lambda x: get_height(join(datapath, x)))\n",
    "modes = data.path.apply(lambda x: get_mode(join(datapath, x)))\n",
    "cls = data.path.apply(lambda x: x.split('/')[1])\n",
    "imgs_info = pd.concat([widths, heights, modes, cls], axis=1, keys=['width', 'height', 'mode', 'cls'])\n",
    "imgs_info.head()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 1440x576 with 10 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(2, 5, figsize=(20, 8))\n",
    "groups = imgs_info.groupby('cls').groups\n",
    "for i, (group_name, group_indxs) in enumerate(groups.items()):\n",
    "    group = imgs_info.loc[group_indxs]\n",
    "    axs[i//5][i%5].scatter(group.width, group.height)\n",
    "    axs[i//5][i%5].set_title(group_name)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "outputs": [
    {
     "data": {
      "text/plain": "           width  height\ncls                     \nn02086240  500.0   375.0\nn02087394  500.0   375.0\nn02088364  500.0   375.0\nn02089973  298.0   250.0\nn02093754  500.0   375.0\nn02096294  500.0   375.0\nn02099601  500.0   375.0\nn02105641  480.0   375.0\nn02111889  500.0   375.0\nn02115641  500.0   375.0",
      "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>width</th>\n      <th>height</th>\n    </tr>\n    <tr>\n      <th>cls</th>\n      <th></th>\n      <th></th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>n02086240</th>\n      <td>500.0</td>\n      <td>375.0</td>\n    </tr>\n    <tr>\n      <th>n02087394</th>\n      <td>500.0</td>\n      <td>375.0</td>\n    </tr>\n    <tr>\n      <th>n02088364</th>\n      <td>500.0</td>\n      <td>375.0</td>\n    </tr>\n    <tr>\n      <th>n02089973</th>\n      <td>298.0</td>\n      <td>250.0</td>\n    </tr>\n    <tr>\n      <th>n02093754</th>\n      <td>500.0</td>\n      <td>375.0</td>\n    </tr>\n    <tr>\n      <th>n02096294</th>\n      <td>500.0</td>\n      <td>375.0</td>\n    </tr>\n    <tr>\n      <th>n02099601</th>\n      <td>500.0</td>\n      <td>375.0</td>\n    </tr>\n    <tr>\n      <th>n02105641</th>\n      <td>480.0</td>\n      <td>375.0</td>\n    </tr>\n    <tr>\n      <th>n02111889</th>\n      <td>500.0</td>\n      <td>375.0</td>\n    </tr>\n    <tr>\n      <th>n02115641</th>\n      <td>500.0</td>\n      <td>375.0</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "imgs_info.groupby('cls').median()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}