Hough_Triplets.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Hough on Triplets\n",
    "\n",
    "Finding valid triplets using a Hough-Transform approach."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from trackml.dataset import load_event\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import random\n",
    "\n",
    "sns.set_style('whitegrid')\n",
    "pd.set_option('display.max_columns', None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "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>hit_id</th>\n",
       "      <th>particle_id</th>\n",
       "      <th>x</th>\n",
       "      <th>y</th>\n",
       "      <th>z</th>\n",
       "      <th>volume_id</th>\n",
       "      <th>layer_id</th>\n",
       "      <th>r</th>\n",
       "      <th>volume_layer</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1507</th>\n",
       "      <td>159</td>\n",
       "      <td>743099092476887040</td>\n",
       "      <td>-42.819302</td>\n",
       "      <td>-39.234200</td>\n",
       "      <td>-1502.5</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>58.075942</td>\n",
       "      <td>7_2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1508</th>\n",
       "      <td>181</td>\n",
       "      <td>743099092476887040</td>\n",
       "      <td>-42.670898</td>\n",
       "      <td>-39.144299</td>\n",
       "      <td>-1498.0</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>57.905800</td>\n",
       "      <td>7_2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1509</th>\n",
       "      <td>1630</td>\n",
       "      <td>743099092476887040</td>\n",
       "      <td>-36.925701</td>\n",
       "      <td>-34.332100</td>\n",
       "      <td>-1302.5</td>\n",
       "      <td>7</td>\n",
       "      <td>4</td>\n",
       "      <td>50.420239</td>\n",
       "      <td>7_4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      hit_id         particle_id          x          y       z  volume_id  \\\n",
       "1507     159  743099092476887040 -42.819302 -39.234200 -1502.5          7   \n",
       "1508     181  743099092476887040 -42.670898 -39.144299 -1498.0          7   \n",
       "1509    1630  743099092476887040 -36.925701 -34.332100 -1302.5          7   \n",
       "\n",
       "      layer_id          r volume_layer  \n",
       "1507         2  58.075942          7_2  \n",
       "1508         2  57.905800          7_2  \n",
       "1509         4  50.420239          7_4  "
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "path2data = '/home/alepfu/Desktop/dataset/train_100_events/'\n",
    "event_id = 'event000001000'\n",
    "hits, cells, particles, truth = load_event(path2data + event_id)\n",
    "\n",
    "data = pd.merge(pd.merge(truth, pd.DataFrame(hits), on = ['hit_id']), pd.DataFrame(particles), on = ['particle_id']).loc[:,['hit_id', 'particle_id', 'x', 'y', 'z','volume_id','layer_id']]\n",
    "data['r'] = data.apply(lambda h: np.sqrt(h['x']**2 + h['y']**2), axis=1)\n",
    "data['volume_layer'] = data.apply(lambda h: str(int(h['volume_id'])) + '_' + str(int(h['layer_id'])), axis=1)\n",
    "\n",
    "N_sample_tracks = 200\n",
    "data_sample = data.loc[data['particle_id'].isin(random.sample(list(data['particle_id'].unique()), N_sample_tracks))].copy()\n",
    "data_sample.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "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>particle_id</th>\n",
       "      <th>volume_layer</th>\n",
       "      <th>hit_id</th>\n",
       "      <th>x</th>\n",
       "      <th>y</th>\n",
       "      <th>z</th>\n",
       "      <th>r</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>22529405569990656</td>\n",
       "      <td>13_8</td>\n",
       "      <td>91750</td>\n",
       "      <td>468.059998</td>\n",
       "      <td>464.967010</td>\n",
       "      <td>-901.599976</td>\n",
       "      <td>659.753349</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>27024002996256770</td>\n",
       "      <td>13_8</td>\n",
       "      <td>94687</td>\n",
       "      <td>417.369995</td>\n",
       "      <td>508.084015</td>\n",
       "      <td>130.000000</td>\n",
       "      <td>657.531048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>27024002996256770</td>\n",
       "      <td>17_2</td>\n",
       "      <td>110915</td>\n",
       "      <td>526.585999</td>\n",
       "      <td>632.679016</td>\n",
       "      <td>179.000000</td>\n",
       "      <td>823.149775</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         particle_id volume_layer  hit_id           x           y           z  \\\n",
       "0  22529405569990656         13_8   91750  468.059998  464.967010 -901.599976   \n",
       "1  27024002996256770         13_8   94687  417.369995  508.084015  130.000000   \n",
       "2  27024002996256770         17_2  110915  526.585999  632.679016  179.000000   \n",
       "\n",
       "            r  \n",
       "0  659.753349  \n",
       "1  657.531048  \n",
       "2  823.149775  "
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get only hits from the centered outer detector strips.\n",
    "hits_outer = data_sample.loc[data_sample['volume_layer'].isin(['17_4','17_2','13_8'])]\n",
    "\n",
    "# Merge hits from overlapping detector panels\n",
    "hits_outer = hits_outer.groupby(['particle_id','volume_layer']).agg({'hit_id':'first', 'x':'mean', 'y':'mean', 'z':'mean', 'r':'mean'}).reset_index()\n",
    "hits_outer.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fbbebc10a58>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(10,5))\n",
    "plt.subplot(1,2,1)\n",
    "plt.plot(hits_outer['z'], hits_outer['r'],\".\")\n",
    "plt.xlabel(\"Z\")\n",
    "plt.ylabel(\"R\")\n",
    "plt.xlim((-3100,3100))\n",
    "plt.ylim((0,1100))\n",
    "plt.subplot(1,2,2)\n",
    "plt.plot(hits_outer['x'], hits_outer['y'],\".\")\n",
    "plt.xlim((-1100,1100))\n",
    "plt.ylim((-1100,1100))\n",
    "plt.xlabel(\"X\")\n",
    "plt.ylabel(\"Y\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "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>hit_id</th>\n",
       "      <th>x</th>\n",
       "      <th>y</th>\n",
       "      <th>z</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>91750</td>\n",
       "      <td>468.059998</td>\n",
       "      <td>464.967010</td>\n",
       "      <td>-901.599976</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>94687</td>\n",
       "      <td>417.369995</td>\n",
       "      <td>508.084015</td>\n",
       "      <td>130.000000</td>\n",
       "      <td>24.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>110915</td>\n",
       "      <td>526.585999</td>\n",
       "      <td>632.679016</td>\n",
       "      <td>179.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   hit_id           x           y           z  label\n",
       "0   91750  468.059998  464.967010 -901.599976    0.0\n",
       "1   94687  417.369995  508.084015  130.000000   24.0\n",
       "2  110915  526.585999  632.679016  179.000000    0.0"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from hough_trans import Clusterer\n",
    "\n",
    "\n",
    "model = Clusterer(N_bins_r0inv=200, N_bins_gamma=500, N_theta=500, min_hits=3) # since we want to form triplets\n",
    "\n",
    "hits_outer = hits_outer.loc[:,['hit_id','x','y', 'z']]\n",
    "labels = model.predict(hits_outer)\n",
    "\n",
    "hits_outer['label'] = labels\n",
    "\n",
    "hits_outer.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fbbecc24eb8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hits_outer['r'] = hits_outer.apply(lambda h: np.sqrt(h['x']**2 + h['y']**2), axis=1)\n",
    "\n",
    "hits_outer_found = hits_outer[hits_outer['label']!=0].copy()\n",
    "\n",
    "fig = plt.figure(figsize=(16,8))\n",
    "plt.subplot(1,2,1)\n",
    "plt.scatter(hits_outer_found['z'], hits_outer_found['r'], c=hits_outer_found['label'], label=hits_outer_found['label'], s=100, cmap='tab20');\n",
    "plt.subplot(1,2,2)\n",
    "plt.scatter(hits_outer_found['x'], hits_outer_found['y'], c=hits_outer_found['label'], label=hits_outer_found['label'], s=100, cmap='tab20');\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accurate found triplets: 0.1111111111111111\n"
     ]
    }
   ],
   "source": [
    "found_triplets = pd.DataFrame(hits_outer[hits_outer['label']!=0].sort_values('label'))\n",
    "check = pd.merge(found_triplets, truth, on=['hit_id'])\n",
    "check = check[['hit_id','particle_id','label']]\n",
    "group = check.groupby('label').agg({'particle_id':pd.Series.nunique})\n",
    "\n",
    "print('Accurate found triplets:', len(group[group['particle_id']==1.0])/len(group))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Triplet finding on the whole dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accurate found triplets: 0.06798328902392708\n",
      "CPU times: user 1min 14s, sys: 769 ms, total: 1min 15s\n",
      "Wall time: 1min 16s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "from hough_trans import Clusterer\n",
    "\n",
    "hits_all = data.copy()\n",
    "\n",
    "# Merge hits from overlapping detector panels\n",
    "hits_all = hits_all.groupby(['particle_id','volume_layer']).agg({'hit_id':'first', 'x':'mean', \n",
    "                                                                 'y':'mean', 'z':'mean', 'r':'mean'}).reset_index()\n",
    "\n",
    "model_all = Clusterer(N_bins_r0inv=200, N_bins_gamma=500, N_theta=500, min_hits=3) # since we want to form triplets\n",
    "hits_all= hits_all.loc[:,['hit_id','x','y', 'z']]\n",
    "labels_all = model_all.predict(hits_all)\n",
    "hits_all['label'] = labels_all\n",
    "\n",
    "found_triplets_all = pd.DataFrame(hits_all[hits_all['label']!=0].sort_values('label'))\n",
    "check_all = pd.merge(found_triplets_all, truth, on=['hit_id'])\n",
    "check_all = check_all[['hit_id','particle_id','label']]\n",
    "group_all = check_all.groupby('label').agg({'particle_id':pd.Series.nunique})\n",
    "\n",
    "print('Accurate found triplets:', len(group_all[group_all['particle_id']==1.0])/len(group_all))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# TODO: proceed here with Hough-space viz"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## TODO: Look how min_hits constraint is set in hough_trans.py"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "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
}