11-dataset-metadata.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.6/importlib/_bootstrap.py:219: RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility. Expected 96, got 88\n",
      "  return f(*args, **kwds)\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sb\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import skimage\n",
    "import skimage.io\n",
    "import skimage.measure\n",
    "\n",
    "import glob"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "MODE = \"TRAINING\" # \"VALIDATION\" | \"TESTING\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "if MODE == \"TRAINING\":\n",
    "    matches = pd.read_csv(\"matched_training_images.csv\")\n",
    "if MODE == \"TESTING\":\n",
    "    matches = pd.read_csv(\"matched_testing_images.csv\")\n",
    "if MODE == \"VALIDATION\":\n",
    "    matches = pd.read_csv(\"matched_validation_images.csv\")\n",
    "\n",
    "matches[\"project\"] = matches.match.str.split(\"/\",  expand=True)[5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f9514ca1d30>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f951b805e80>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sb.barplot(data=matches.groupby(\"project\").count().reset_index(), x=\"project\", y=\"match\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "if MODE == \"TRAINING\":\n",
    "    gt_dir = \"/data1/image-segmentation/dsb_eval/raw_annotations/\"\n",
    "elif MODE == \"TESTING\":\n",
    "    gt_dir = \"/storage/data/DSB2018/phase-2-gt/\"\n",
    "elif MODE == \"VALIDATION\":\n",
    "    gt_dir = \"/storage/data/DSB2018/phase-1-gt/\"\n",
    "\n",
    "matches[\"match_width\"] = 0\n",
    "matches[\"match_height\"] = 0\n",
    "matches[\"crop_width\"] = 0\n",
    "matches[\"crop_height\"] = 0\n",
    "matches[\"SNR\"] = 0\n",
    "matches[\"CR\"] = 0\n",
    "matches[\"objects\"] = 0\n",
    "matches[\"object_size\"] = 0\n",
    "\n",
    "for k,r in matches.iterrows():\n",
    "    # Handle weird file names\n",
    "    if MODE == \"TESTING\":\n",
    "        r.match = r.match.encode(\"utf-16\", \"strict\").decode(\"utf-16\")\n",
    "        s = r.match.encode(\"utf-16\").decode(\"utf-16\").replace(\"ó\",\"**\")\n",
    "        s = r.match.encode(\"utf-16\").decode(\"utf-16\").replace(\"β\",\"**\")\n",
    "        files = glob.glob(s)\n",
    "        r.match = files[0]\n",
    "\n",
    "    # Dimensions of original image\n",
    "    original = skimage.io.imread(r.match)\n",
    "    matches.loc[k,\"match_width\"] = original.shape[0]\n",
    "    matches.loc[k,\"match_height\"] = original.shape[1]\n",
    "    \n",
    "    # Dimensions of the crop\n",
    "    crop = skimage.io.imread(r.crop)\n",
    "    matches.loc[k,\"crop_width\"] = crop.shape[0]\n",
    "    matches.loc[k,\"crop_height\"] = crop.shape[1]\n",
    "    \n",
    "    # Load ground truth\n",
    "    if MODE == \"TRAINING\":\n",
    "        masks = skimage.io.imread(gt_dir + r.crop.split(\"/\")[-1].replace(\".png\", \".tiff\"))\n",
    "    elif MODE == \"TESTING\":\n",
    "        masks = skimage.io.imread(gt_dir + r.crop.split(\"/\")[-1])\n",
    "    elif MODE == \"VALIDATION\":\n",
    "        masks = skimage.io.imread(gt_dir + r.crop.split(\"/\")[-1].replace(\".png\", \".tiff\"))\n",
    "    if len(crop.shape) > 2:\n",
    "        crop = np.mean(crop, axis=2)\n",
    "    matches.loc[k,\"objects\"] = len(np.unique(masks)) - 1\n",
    "        \n",
    "    # SNR\n",
    "    bg_avg = np.mean( (masks == 0)*crop)\n",
    "    fg_avg = np.mean( (masks > 0)*crop)\n",
    "    bg_std = np.std( (masks == 0)*crop)\n",
    "    matches.loc[k, \"SNR\"] = np.abs(fg_avg - bg_avg) / bg_std\n",
    "    \n",
    "    # Contrast Ratio\n",
    "    matches.loc[k, \"CR\"] = fg_avg / bg_avg\n",
    "    \n",
    "    # Average object size\n",
    "    area = 0\n",
    "    regions = skimage.measure.regionprops(masks)\n",
    "    for region in regions:\n",
    "        area += region.area\n",
    "    matches.loc[k, \"object_size\"] = area/len(regions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'/storage/data/DSB2018/image-sets/bbbc018/bbbc018-human-ht29-colon-cancer-cells_15961-DNA.png'"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r.match.encode(\"utf-16\").decode(\"utf-16\")"
   ]
  },
  {
   "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>Unnamed: 0</th>\n",
       "      <th>coord_x</th>\n",
       "      <th>coord_y</th>\n",
       "      <th>exact</th>\n",
       "      <th>match_width</th>\n",
       "      <th>match_height</th>\n",
       "      <th>crop_width</th>\n",
       "      <th>crop_height</th>\n",
       "      <th>SNR</th>\n",
       "      <th>CR</th>\n",
       "      <th>objects</th>\n",
       "      <th>object_size</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>project</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>bbbc002</th>\n",
       "      <td>364.277778</td>\n",
       "      <td>128.000000</td>\n",
       "      <td>128.000000</td>\n",
       "      <td>True</td>\n",
       "      <td>512.0</td>\n",
       "      <td>512.0</td>\n",
       "      <td>256.0</td>\n",
       "      <td>256.0</td>\n",
       "      <td>1.615410</td>\n",
       "      <td>0.340140</td>\n",
       "      <td>40.930556</td>\n",
       "      <td>306.247659</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bbbc018</th>\n",
       "      <td>317.458333</td>\n",
       "      <td>128.000000</td>\n",
       "      <td>128.000000</td>\n",
       "      <td>True</td>\n",
       "      <td>512.0</td>\n",
       "      <td>512.0</td>\n",
       "      <td>256.0</td>\n",
       "      <td>256.0</td>\n",
       "      <td>4.812093</td>\n",
       "      <td>0.092777</td>\n",
       "      <td>30.611111</td>\n",
       "      <td>126.136061</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bbbc022</th>\n",
       "      <td>326.116279</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>True</td>\n",
       "      <td>520.0</td>\n",
       "      <td>696.0</td>\n",
       "      <td>520.0</td>\n",
       "      <td>696.0</td>\n",
       "      <td>1.646974</td>\n",
       "      <td>0.347120</td>\n",
       "      <td>115.813953</td>\n",
       "      <td>623.221590</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_06Eggert</th>\n",
       "      <td>340.938776</td>\n",
       "      <td>254.693878</td>\n",
       "      <td>355.102041</td>\n",
       "      <td>True</td>\n",
       "      <td>1040.0</td>\n",
       "      <td>1392.0</td>\n",
       "      <td>520.0</td>\n",
       "      <td>696.0</td>\n",
       "      <td>2.415012</td>\n",
       "      <td>0.294603</td>\n",
       "      <td>93.183673</td>\n",
       "      <td>1059.452655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_07mito</th>\n",
       "      <td>404.000000</td>\n",
       "      <td>104.000000</td>\n",
       "      <td>138.800000</td>\n",
       "      <td>True</td>\n",
       "      <td>520.0</td>\n",
       "      <td>694.0</td>\n",
       "      <td>260.0</td>\n",
       "      <td>347.0</td>\n",
       "      <td>1.689949</td>\n",
       "      <td>0.315849</td>\n",
       "      <td>81.600000</td>\n",
       "      <td>142.241824</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_07mito_fibro</th>\n",
       "      <td>430.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>True</td>\n",
       "      <td>1040.0</td>\n",
       "      <td>1388.0</td>\n",
       "      <td>1040.0</td>\n",
       "      <td>1388.0</td>\n",
       "      <td>2.948662</td>\n",
       "      <td>0.180198</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>7244.071429</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_14ISBI</th>\n",
       "      <td>301.750000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>True</td>\n",
       "      <td>1024.0</td>\n",
       "      <td>1024.0</td>\n",
       "      <td>1024.0</td>\n",
       "      <td>1024.0</td>\n",
       "      <td>5.981610</td>\n",
       "      <td>0.011893</td>\n",
       "      <td>84.062500</td>\n",
       "      <td>242.483783</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_15Stachler_bw</th>\n",
       "      <td>398.500000</td>\n",
       "      <td>128.000000</td>\n",
       "      <td>160.000000</td>\n",
       "      <td>True</td>\n",
       "      <td>512.0</td>\n",
       "      <td>640.0</td>\n",
       "      <td>256.0</td>\n",
       "      <td>320.0</td>\n",
       "      <td>5.797694</td>\n",
       "      <td>0.382979</td>\n",
       "      <td>26.750000</td>\n",
       "      <td>384.833072</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_15Stachler_c1</th>\n",
       "      <td>313.969697</td>\n",
       "      <td>124.121212</td>\n",
       "      <td>160.000000</td>\n",
       "      <td>True</td>\n",
       "      <td>512.0</td>\n",
       "      <td>640.0</td>\n",
       "      <td>256.0</td>\n",
       "      <td>320.0</td>\n",
       "      <td>2.131907</td>\n",
       "      <td>0.242167</td>\n",
       "      <td>39.439394</td>\n",
       "      <td>723.135541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_15Stachler_c2</th>\n",
       "      <td>388.000000</td>\n",
       "      <td>128.000000</td>\n",
       "      <td>160.000000</td>\n",
       "      <td>True</td>\n",
       "      <td>512.0</td>\n",
       "      <td>640.0</td>\n",
       "      <td>256.0</td>\n",
       "      <td>320.0</td>\n",
       "      <td>1.883664</td>\n",
       "      <td>0.162996</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>1085.400183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_15jager</th>\n",
       "      <td>297.461538</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>True</td>\n",
       "      <td>512.0</td>\n",
       "      <td>640.0</td>\n",
       "      <td>512.0</td>\n",
       "      <td>640.0</td>\n",
       "      <td>2.411154</td>\n",
       "      <td>0.167754</td>\n",
       "      <td>31.615385</td>\n",
       "      <td>1243.527609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_16ure_human</th>\n",
       "      <td>309.000000</td>\n",
       "      <td>128.000000</td>\n",
       "      <td>128.000000</td>\n",
       "      <td>True</td>\n",
       "      <td>512.0</td>\n",
       "      <td>512.0</td>\n",
       "      <td>256.0</td>\n",
       "      <td>256.0</td>\n",
       "      <td>6.198831</td>\n",
       "      <td>0.038206</td>\n",
       "      <td>10.511905</td>\n",
       "      <td>158.771348</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_16ure_pig</th>\n",
       "      <td>352.794118</td>\n",
       "      <td>128.000000</td>\n",
       "      <td>135.529412</td>\n",
       "      <td>True</td>\n",
       "      <td>512.0</td>\n",
       "      <td>512.0</td>\n",
       "      <td>256.0</td>\n",
       "      <td>256.0</td>\n",
       "      <td>3.749859</td>\n",
       "      <td>0.134992</td>\n",
       "      <td>38.352941</td>\n",
       "      <td>131.455713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_17Keller</th>\n",
       "      <td>106.833333</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>True</td>\n",
       "      <td>603.0</td>\n",
       "      <td>1272.0</td>\n",
       "      <td>603.0</td>\n",
       "      <td>1272.0</td>\n",
       "      <td>4.425306</td>\n",
       "      <td>0.070501</td>\n",
       "      <td>229.166667</td>\n",
       "      <td>166.189624</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_17Marinaccio</th>\n",
       "      <td>400.289474</td>\n",
       "      <td>121.263158</td>\n",
       "      <td>160.000000</td>\n",
       "      <td>True</td>\n",
       "      <td>512.0</td>\n",
       "      <td>640.0</td>\n",
       "      <td>256.0</td>\n",
       "      <td>320.0</td>\n",
       "      <td>1.694530</td>\n",
       "      <td>0.222714</td>\n",
       "      <td>50.736842</td>\n",
       "      <td>364.833832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_17Moscow</th>\n",
       "      <td>358.549451</td>\n",
       "      <td>181.978022</td>\n",
       "      <td>178.021978</td>\n",
       "      <td>True</td>\n",
       "      <td>720.0</td>\n",
       "      <td>720.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>0.911741</td>\n",
       "      <td>0.461562</td>\n",
       "      <td>23.483516</td>\n",
       "      <td>1561.249397</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Unnamed: 0     coord_x     coord_y  exact  match_width  \\\n",
       "project                                                                   \n",
       "bbbc002          364.277778  128.000000  128.000000   True        512.0   \n",
       "bbbc018          317.458333  128.000000  128.000000   True        512.0   \n",
       "bbbc022          326.116279    0.000000    0.000000   True        520.0   \n",
       "p_06Eggert       340.938776  254.693878  355.102041   True       1040.0   \n",
       "p_07mito         404.000000  104.000000  138.800000   True        520.0   \n",
       "p_07mito_fibro   430.000000    0.000000    0.000000   True       1040.0   \n",
       "p_14ISBI         301.750000    0.000000    0.000000   True       1024.0   \n",
       "p_15Stachler_bw  398.500000  128.000000  160.000000   True        512.0   \n",
       "p_15Stachler_c1  313.969697  124.121212  160.000000   True        512.0   \n",
       "p_15Stachler_c2  388.000000  128.000000  160.000000   True        512.0   \n",
       "p_15jager        297.461538    0.000000    0.000000   True        512.0   \n",
       "p_16ure_human    309.000000  128.000000  128.000000   True        512.0   \n",
       "p_16ure_pig      352.794118  128.000000  135.529412   True        512.0   \n",
       "p_17Keller       106.833333    0.000000    0.000000   True        603.0   \n",
       "p_17Marinaccio   400.289474  121.263158  160.000000   True        512.0   \n",
       "p_17Moscow       358.549451  181.978022  178.021978   True        720.0   \n",
       "\n",
       "                 match_height  crop_width  crop_height       SNR        CR  \\\n",
       "project                                                                      \n",
       "bbbc002                 512.0       256.0        256.0  1.615410  0.340140   \n",
       "bbbc018                 512.0       256.0        256.0  4.812093  0.092777   \n",
       "bbbc022                 696.0       520.0        696.0  1.646974  0.347120   \n",
       "p_06Eggert             1392.0       520.0        696.0  2.415012  0.294603   \n",
       "p_07mito                694.0       260.0        347.0  1.689949  0.315849   \n",
       "p_07mito_fibro         1388.0      1040.0       1388.0  2.948662  0.180198   \n",
       "p_14ISBI               1024.0      1024.0       1024.0  5.981610  0.011893   \n",
       "p_15Stachler_bw         640.0       256.0        320.0  5.797694  0.382979   \n",
       "p_15Stachler_c1         640.0       256.0        320.0  2.131907  0.242167   \n",
       "p_15Stachler_c2         640.0       256.0        320.0  1.883664  0.162996   \n",
       "p_15jager               640.0       512.0        640.0  2.411154  0.167754   \n",
       "p_16ure_human           512.0       256.0        256.0  6.198831  0.038206   \n",
       "p_16ure_pig             512.0       256.0        256.0  3.749859  0.134992   \n",
       "p_17Keller             1272.0       603.0       1272.0  4.425306  0.070501   \n",
       "p_17Marinaccio          640.0       256.0        320.0  1.694530  0.222714   \n",
       "p_17Moscow              720.0       360.0        360.0  0.911741  0.461562   \n",
       "\n",
       "                    objects  object_size  \n",
       "project                                   \n",
       "bbbc002           40.930556   306.247659  \n",
       "bbbc018           30.611111   126.136061  \n",
       "bbbc022          115.813953   623.221590  \n",
       "p_06Eggert        93.183673  1059.452655  \n",
       "p_07mito          81.600000   142.241824  \n",
       "p_07mito_fibro    14.000000  7244.071429  \n",
       "p_14ISBI          84.062500   242.483783  \n",
       "p_15Stachler_bw   26.750000   384.833072  \n",
       "p_15Stachler_c1   39.439394   723.135541  \n",
       "p_15Stachler_c2   12.000000  1085.400183  \n",
       "p_15jager         31.615385  1243.527609  \n",
       "p_16ure_human     10.511905   158.771348  \n",
       "p_16ure_pig       38.352941   131.455713  \n",
       "p_17Keller       229.166667   166.189624  \n",
       "p_17Marinaccio    50.736842   364.833832  \n",
       "p_17Moscow        23.483516  1561.249397  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matches.groupby(\"project\").mean()#[\"coord_x\",\"coord_y\", \"match_width\", \"match_height\", \"crop_width\", \"crop_height\", \"SNR\", \"CR\"].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>exact</th>\n",
       "      <th>objects</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>project</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>bbbc002</th>\n",
       "      <td>72.0</td>\n",
       "      <td>2947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bbbc018</th>\n",
       "      <td>144.0</td>\n",
       "      <td>4408</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bbbc022</th>\n",
       "      <td>43.0</td>\n",
       "      <td>4980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_06Eggert</th>\n",
       "      <td>49.0</td>\n",
       "      <td>4566</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_07mito</th>\n",
       "      <td>5.0</td>\n",
       "      <td>408</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_07mito_fibro</th>\n",
       "      <td>1.0</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_14ISBI</th>\n",
       "      <td>16.0</td>\n",
       "      <td>1345</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_15Stachler_bw</th>\n",
       "      <td>4.0</td>\n",
       "      <td>107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_15Stachler_c1</th>\n",
       "      <td>66.0</td>\n",
       "      <td>2603</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_15Stachler_c2</th>\n",
       "      <td>4.0</td>\n",
       "      <td>48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_15jager</th>\n",
       "      <td>13.0</td>\n",
       "      <td>411</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_16ure_human</th>\n",
       "      <td>84.0</td>\n",
       "      <td>883</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_16ure_pig</th>\n",
       "      <td>34.0</td>\n",
       "      <td>1304</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_17Keller</th>\n",
       "      <td>6.0</td>\n",
       "      <td>1375</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_17Marinaccio</th>\n",
       "      <td>38.0</td>\n",
       "      <td>1928</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p_17Moscow</th>\n",
       "      <td>91.0</td>\n",
       "      <td>2137</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 exact  objects\n",
       "project                        \n",
       "bbbc002           72.0     2947\n",
       "bbbc018          144.0     4408\n",
       "bbbc022           43.0     4980\n",
       "p_06Eggert        49.0     4566\n",
       "p_07mito           5.0      408\n",
       "p_07mito_fibro     1.0       14\n",
       "p_14ISBI          16.0     1345\n",
       "p_15Stachler_bw    4.0      107\n",
       "p_15Stachler_c1   66.0     2603\n",
       "p_15Stachler_c2    4.0       48\n",
       "p_15jager         13.0      411\n",
       "p_16ure_human     84.0      883\n",
       "p_16ure_pig       34.0     1304\n",
       "p_17Keller         6.0     1375\n",
       "p_17Marinaccio    38.0     1928\n",
       "p_17Moscow        91.0     2137"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matches.groupby(\"project\")[\"exact\",\"objects\"].sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x7f9514c81ba8>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f9514c81080>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sb.lmplot(data=matches.groupby(\"project\").mean().reset_index(), x=\"SNR\", y=\"CR\", fit_reg=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "if MODE == \"TESTING\":\n",
    "    test_meta = pd.read_csv(\"/storage/data/DSB2018/metadata/dsb_summary_metadata_stage2testdata_v2.csv\")\n",
    "    test_groups = pd.read_csv(\"/storage/data/DSB2018/metadata/image_groups_defined_by_beth.csv\")\n",
    "\n",
    "    matches[\"hash_name\"] = matches[\"crop\"].str.split(\"/\", expand=True)[5].str.split(\".\", expand=True)[0]\n",
    "    full_meta = pd.merge(test_meta, matches, left_on=\"hash_name\", right_on=\"hash_name\").reset_index()\n",
    "    full_meta = pd.merge(full_meta, test_groups, left_on=\"hash_name\", right_on=\"Image\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "if MODE == \"TESTING\":\n",
    "    x = full_meta.groupby(\"project_id\")[\"SNR\", \"CR\", \"crop_width\", \"crop_height\", \"match_width\", \"match_height\", \"object_size\"].mean()\n",
    "    # Manually print x below"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "if MODE == \"TESTING\":\n",
    "    y = full_meta.groupby(\"project_id\")[\"index\", \"objects\"].sum()\n",
    "    # Manually print y below"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}