PokemonGo_RegionComparisons.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# How do I make this work?\n",
    "\n",
    "Possibilities:\n",
    " - I'm viewing this in github and I just want to preview the end results\n",
    "   - View this notebook with nbviewer (follow [this link](https://nbviewer.jupyter.org/gist/egentry/e212c944f3d127fb96a25b885aa35e3e)), then scroll to the bottom. The table should be sortable. I was hoping you could view the table directly in Github, but it wasn't working.\n",
    "   \n",
    "   \n",
    "   \n",
    "   \n",
    " - I want to run this on my own datasets, but don't want to change much else.\n",
    " \n",
    "   1) Open in a `jupyter notebook` or just copy the code cells into a python script.\n",
    "   \n",
    "   2a) If using a `pokeminer`-style database: change the `databases` list [below](#Chose-what-datasets-to-compare), then hit `Cell > Run All` in the notebook menu at the top\n",
    "\n",
    "   2b) If using some different dataset: go to [this spot](#Create-a-Data-object), and create the `Data` object and `db_labels` yourself. Remember `data.counts` should be a 2D array, of shape (N_POKEMON X 2)\n",
    "   \n",
    " - I like the idea, but want to change a lot.\n",
    "   - This is all MIT licensed, so you can do what you want.  I might be able to offer some help troubleshooting / extending this, if you're interested. \n",
    " \n",
    "\n",
    "\n",
    "\n",
    "## Misc. details\n",
    "\n",
    "Works in Python 2 and 3\n",
    "\n",
    "Required packages:\n",
    " - numpy\n",
    " - pandas\n",
    " - scipy\n",
    " - astropy\n",
    " - ipython + jupyter notebook (if you want to run this in a notebook rather than a python script)\n",
    "\n",
    "All of those packages are included in the default anaconda distribution of python, or though pip\n",
    "\n",
    "License: MIT"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from __future__ import print_function, division\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "from scipy import stats\n",
    "from scipy.special import gamma, gammaln\n",
    "\n",
    "#if you know a better way to push a table to a notebook, let me know\n",
    "from astropy.table import Table\n",
    "\n",
    "N_POKEMON = 151"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Chose what datasets to compare"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# a list of tuples of the form (\"db_uri\", \"Dataset Label\")\n",
    "databases = [\n",
    "    (\"sqlite:///db_Paris.sqlite\", \"Paris\"),\n",
    "    (\"sqlite:///db_SC.sqlite\", \"Santa Cruz\"),\n",
    "]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Create a Data object\n",
    "\n",
    "This can either be from a list of sqllite databases (formatted in the `pokeminer` style) or you can just create a `Data` object by hand, by passing `Data` a counts array of shape `N_POKEMON x N_DATASETS`\n",
    "\n",
    "I currently assume you're only comparing 2 datasets, but it wouldn't be that hard to add more, if you're willing to play with code."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def read_sql_into_counts_array(db_uri):\n",
    "    \"\"\"Given a database URI\n",
    "    Inputs\n",
    "    ------\n",
    "        db_uri : database connectable recognized by pandas.read_sql_table\n",
    "        \n",
    "    Returns\n",
    "    -------\n",
    "        counts : numpy.array\n",
    "            - shape: (151,)\n",
    "            - dtype: int\n",
    "        \n",
    "    Notes\n",
    "    -----\n",
    "        Expects db_uri to have a table `sightings` which has the \n",
    "        columns defined in `db.py` of `pokeminer`\n",
    "    \"\"\"\n",
    "    df = pd.read_sql_table(\"sightings\", db_uri)\n",
    "\n",
    "    counts = np.zeros(151, dtype=int)\n",
    "    for p_id in df.pokemon_id:\n",
    "        counts[p_id-1] += 1\n",
    "    return counts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "class Data(object):\n",
    "    def __init__(self, counts):\n",
    "        self.counts = counts\n",
    "        \n",
    "    @staticmethod\n",
    "    def from_db_uris(db_uris):\n",
    "        \"\"\"Container for observed data\n",
    "        \n",
    "        Inputs\n",
    "        ------\n",
    "            db_uris: sequence of `db_uri` for `read_sql_into_counts_array`\n",
    "                - must be of length `n_regions`\n",
    "        \"\"\"\n",
    "        counts = np.empty((N_POKEMON, len(db_uris)),\n",
    "                          dtype=int)\n",
    "        \n",
    "        for i, db_uri in enumerate(db_uris):\n",
    "            counts[:,i] = read_sql_into_counts_array(db_uri)\n",
    "        \n",
    "        return Data(counts)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "n_regions = len(databases)\n",
    "\n",
    "db_uris = [db_tuple[0] for db_tuple in databases]\n",
    "db_labels = [db_tuple[1] for db_tuple in databases]\n",
    "\n",
    "data = Data.from_db_uris(db_uris)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model comparisons\n",
    "\n",
    "We're about to do some statistics to compare two different models, to see which describes our data.  In short, Model 1 says that spawn rates of a particular species can be different between two regions. Model 2 says the spawn rate is the same in both regions.\n",
    "\n",
    "\n",
    "## Technical details\n",
    "\n",
    "For each Pokemon species, with Pokemon number $j^\\prime$, we have $N$ total counts between the two regions. We assume the distribution between these two regions follows a [multinomial distribution](https://en.wikipedia.org/wiki/Multinomial_distribution). (This also happens to be a binomial distribution, but keeping it as a multinomial makes it easier to add more regions.)\n",
    "\n",
    "The difference between my models comes from the Bayesian prior.  Both models use a [Dirichlet prior](https://en.wikipedia.org/wiki/Dirichlet_distribution), but they differ in their hyperparameters, $\\alpha_i$. \n",
    "\n",
    "\"Model 1\" uses a non-informative prior $\\alpha_i = 1/2$. When we want to estimate the relative spawn ratio, this is effectively like adding an extra $1/2$ to the number of observations in each region. (Because [reasons](https://en.wikipedia.org/wiki/Jeffreys_prior)) \n",
    "\n",
    "\"Model 2\" uses a highly-informative prior.  If we're only looking at 1 Pokemon species at a time, then we assume its spawn ratio will be _very_ similar to the combined spawn ratio of all Pokemon except this particular species. For a region $i$ and Pokemon number $j^\\prime$ this gives: $\\alpha_i = 1/2 + \\sum_{j \\neq j^\\prime} \\mathtt{counts[i,j]}$\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def log_B(alpha):\n",
    "    \"\"\"Calculates the natural log of B, a function defined by:\n",
    "    https://en.wikipedia.org/wiki/Dirichlet_distribution\n",
    "    \n",
    "    Inputs\n",
    "    ------\n",
    "    alpha : numpy.ndarray\n",
    "        - an array of the alpha_i parameters of a Dirichlet distribution\n",
    "        \n",
    "    Returns\n",
    "    -------\n",
    "    log_B_ : float\n",
    "    \n",
    "    Notes\n",
    "    -----\n",
    "    alpha_i > 0, for all alpha_i\n",
    "    \"\"\"\n",
    "    log_B_ = np.sum(gammaln(alpha)) - gammaln(np.sum(alpha))\n",
    "    return log_B_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def log_marginal(prior, counts):\n",
    "    \"\"\"Calculate the natural log of the marginal density for the observed counts.\n",
    "    \n",
    "    Inputs\n",
    "    ------\n",
    "        - prior : scipy.stats.Dirichlet object\n",
    "            - the `a priori` distribution\n",
    "        - counts : numpy.ndarray\n",
    "            - the number of observed pokemon for each category\n",
    "            - must be flattened into a 1D array\n",
    "    \n",
    "    Returns\n",
    "    -------\n",
    "        - log_marginal_ : float\n",
    "    \n",
    "    Notes\n",
    "    -----\n",
    "        - This is also known as the `prior predictive`\n",
    "        - This calculates:\n",
    "            ln p(x) = ln \\int( p(x|theta) * p(theta)  d theta)\n",
    "            where\n",
    "            p(theta) is the prior (a dirichlet distribution)\n",
    "            and\n",
    "            p(x|theta) is the likelihood (a multinomial distribution)\n",
    "    \"\"\"\n",
    "    n = counts.sum()\n",
    "    alpha = prior.alpha\n",
    "    beta = alpha + counts\n",
    "\n",
    "    log_B_ratio = log_B(beta) - log_B(alpha)\n",
    "\n",
    "    log_factorial_ratio = gammaln(n+1) - np.sum(gammaln(counts+1))\n",
    "\n",
    "    log_marginal_ = log_B_ratio + log_factorial_ratio\n",
    "    return log_marginal_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def log_bayes_factor(prior_1, prior_2, counts):\n",
    "    \"\"\"Calculate the natural log of the Bayes factor between two hypotheses.\n",
    "    \n",
    "    Inputs\n",
    "    ------\n",
    "        - prior_1 : scipy.stats.Dirichlet object\n",
    "            - the `a priori` distribution of hypothesis 1\n",
    "        - prior_2 : scipy.stats.Dirichlet object\n",
    "            - the `a priori` distribution of hypothesis 2\n",
    "        - counts : numpy.ndarray\n",
    "            - the number of observed pokemon for each category\n",
    "            - must be flattened into a 1D array\n",
    "    \n",
    "    Returns\n",
    "    -------\n",
    "        - log_bayes_factor_ : float\n",
    "    \n",
    "    Notes\n",
    "    -----\n",
    "        - This calculates:\n",
    "            ln BF = ln (p(x|model 1) / p(x|model 2))\n",
    "        - If BF > 0, the data favors model 1;\n",
    "          If BF < 0, the data favors model 2;\n",
    "        - You need abs( ln BF ) > 1 to have any confidence\n",
    "          You need abs( ln BF ) > 5 to be very confident\n",
    "        - For more information see:\n",
    "          https://en.wikipedia.org/wiki/Bayes_factor\n",
    "            \n",
    "    \"\"\"\n",
    "    log_bayes_factor_ = log_marginal(prior_1, counts) - log_marginal(prior_2, counts)\n",
    "    return log_bayes_factor_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def region_dependent_label(log_BF):\n",
    "    if   log_BF < -5:\n",
    "        return \"No - very confident\"\n",
    "    elif log_BF < -1:\n",
    "        return \"No\"\n",
    "    elif log_BF < 1:\n",
    "        return \"Not sure\"\n",
    "    elif log_BF < 5:\n",
    "        return \"Yes\"\n",
    "    else:\n",
    "        return \"Yes - very confident\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "prior_overall = stats.dirichlet(np.full(data.counts.size, 1/2))\n",
    "\n",
    "prior_1 = stats.dirichlet(np.full(n_regions, 1/2))\n",
    "\n",
    "log_bayes_factors = np.empty(N_POKEMON)\n",
    "for j in range(N_POKEMON):\n",
    "    counts_j = data.counts[j]\n",
    "\n",
    "    mask = np.arange(N_POKEMON) != j # select all pokemon *except* this one\n",
    "    counts_except_j = data.counts[mask].sum(axis=0)\n",
    "\n",
    "    prior_2_j = stats.dirichlet(counts_except_j + 1/2)\n",
    "    \n",
    "    log_bayes_factors[j] = log_bayes_factor(\n",
    "        prior_1, \n",
    "        prior_2_j, \n",
    "        counts_j,\n",
    "    )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Gather together some useful results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Taken from `locales/pokemon.en.json` from pokeminer\n",
    "\n",
    "pokemon_name_map = {1: 'Bulbasaur', 2: 'Ivysaur', 3: 'Venusaur', 4: 'Charmander', 5: 'Charmeleon', 6: 'Charizard', 7: 'Squirtle', 8: 'Wartortle', 9: 'Blastoise', 10: 'Caterpie', 11: 'Metapod', 12: 'Butterfree', 13: 'Weedle', 14: 'Kakuna', 15: 'Beedrill', 16: 'Pidgey', 17: 'Pidgeotto', 18: 'Pidgeot', 19: 'Rattata', 20: 'Raticate', 21: 'Spearow', 22: 'Fearow', 23: 'Ekans', 24: 'Arbok', 25: 'Pikachu', 26: 'Raichu', 27: 'Sandshrew', 28: 'Sandslash', 29: 'Nidoran(F)', 30: 'Nidorina', 31: 'Nidoqueen', 32: 'Nidoran(M)', 33: 'Nidorino', 34: 'Nidoking', 35: 'Clefairy', 36: 'Clefable', 37: 'Vulpix', 38: 'Ninetales', 39: 'Jigglypuff', 40: 'Wigglytuff', 41: 'Zubat', 42: 'Golbat', 43: 'Oddish', 44: 'Gloom', 45: 'Vileplume', 46: 'Paras', 47: 'Parasect', 48: 'Venonat', 49: 'Venomoth', 50: 'Diglett', 51: 'Dugtrio', 52: 'Meowth', 53: 'Persian', 54: 'Psyduck', 55: 'Golduck', 56: 'Mankey', 57: 'Primeape', 58: 'Growlithe', 59: 'Arcanine', 60: 'Poliwag', 61: 'Poliwhirl', 62: 'Poliwrath', 63: 'Abra', 64: 'Kadabra', 65: 'Alakazam', 66: 'Machop', 67: 'Machoke', 68: 'Machamp', 69: 'Bellsprout', 70: 'Weepinbell', 71: 'Victreebel', 72: 'Tentacool', 73: 'Tentacruel', 74: 'Geodude', 75: 'Graveler', 76: 'Golem', 77: 'Ponyta', 78: 'Rapidash', 79: 'Slowpoke', 80: 'Slowbro', 81: 'Magnemite', 82: 'Magneton', 83: \"Farfetch'd\", 84: 'Doduo', 85: 'Dodrio', 86: 'Seel', 87: 'Dewgong', 88: 'Grimer', 89: 'Muk', 90: 'Shellder', 91: 'Cloyster', 92: 'Gastly', 93: 'Haunter', 94: 'Gengar', 95: 'Onix', 96: 'Drowzee', 97: 'Hypno', 98: 'Krabby', 99: 'Kingler', 100: 'Voltorb', 101: 'Electrode', 102: 'Exeggcute', 103: 'Exeggutor', 104: 'Cubone', 105: 'Marowak', 106: 'Hitmonlee', 107: 'Hitmonchan', 108: 'Lickitung', 109: 'Koffing', 110: 'Weezing', 111: 'Rhyhorn', 112: 'Rhydon', 113: 'Chansey', 114: 'Tangela', 115: 'Kangaskhan', 116: 'Horsea', 117: 'Seadra', 118: 'Goldeen', 119: 'Seaking', 120: 'Staryu', 121: 'Starmie', 122: 'Mr. Mime', 123: 'Scyther', 124: 'Jynx', 125: 'Electabuzz', 126: 'Magmar', 127: 'Pinsir', 128: 'Tauros', 129: 'Magikarp', 130: 'Gyarados', 131: 'Lapras', 132: 'Ditto', 133: 'Eevee', 134: 'Vaporeon', 135: 'Jolteon', 136: 'Flareon', 137: 'Porygon', 138: 'Omanyte', 139: 'Omastar', 140: 'Kabuto', 141: 'Kabutops', 142: 'Aerodactyl', 143: 'Snorlax', 144: 'Articuno', 145: 'Zapdos', 146: 'Moltres', 147: 'Dratini', 148: 'Dragonair', 149: 'Dragonite', 150: 'Mewtwo', 151: 'Mew', 152: 'Chikorita', 153: 'Bayleef', 154: 'Meganium', 155: 'Cyndaquil', 156: 'Quilava', 157: 'Typhlosion', 158: 'Totodile', 159: 'Croconaw', 160: 'Feraligatr', 161: 'Sentret', 162: 'Furret', 163: 'Hoothoot', 164: 'Noctowl', 165: 'Ledyba', 166: 'Ledian', 167: 'Spinarak', 168: 'Ariados', 169: 'Crobat', 170: 'Chinchou', 171: 'Lanturn', 172: 'Pichu', 173: 'Cleffa', 174: 'Igglybuff', 175: 'Togepi', 176: 'Togetic', 177: 'Natu', 178: 'Xatu', 179: 'Mareep', 180: 'Flaaffy', 181: 'Ampharos', 182: 'Bellossom', 183: 'Marill', 184: 'Azumarill', 185: 'Sudowoodo', 186: 'Politoed', 187: 'Hoppip', 188: 'Skiploom', 189: 'Jumpluff', 190: 'Aipom', 191: 'Sunkern', 192: 'Sunflora', 193: 'Yanma', 194: 'Wooper', 195: 'Quagsire', 196: 'Espeon', 197: 'Umbreon', 198: 'Murkrow', 199: 'Slowking', 200: 'Misdreavus', 201: 'Unown', 202: 'Wobbuffet', 203: 'Girafarig', 204: 'Pineco', 205: 'Forretress', 206: 'Dunsparce', 207: 'Gligar', 208: 'Steelix', 209: 'Snubbull', 210: 'Granbull', 211: 'Qwilfish', 212: 'Scizor', 213: 'Shuckle', 214: 'Heracross', 215: 'Sneasel', 216: 'Teddiursa', 217: 'Ursaring', 218: 'Slugma', 219: 'Magcargo', 220: 'Swinub', 221: 'Piloswine', 222: 'Corsola', 223: 'Remoraid', 224: 'Octillery', 225: 'Delibird', 226: 'Mantine', 227: 'Skarmory', 228: 'Houndour', 229: 'Houndoom', 230: 'Kingdra', 231: 'Phanpy', 232: 'Donphan', 233: 'Porygon2', 234: 'Stantler', 235: 'Smeargle', 236: 'Tyrogue', 237: 'Hitmontop', 238: 'Smoochum', 239: 'Elekid', 240: 'Magby', 241: 'Miltank', 242: 'Blissey', 243: 'Raikou', 244: 'Entei', 245: 'Suicune', 246: 'Larvitar', 247: 'Pupitar', 248: 'Tyranitar', 249: 'Lugia', 250: 'Ho-Oh', 251: 'Celebi', 252: 'Treecko', 253: 'Grovyle', 254: 'Sceptile', 255: 'Torchic', 256: 'Combusken', 257: 'Blaziken', 258: 'Mudkip', 259: 'Marshtomp', 260: 'Swampert', 261: 'Poochyena', 262: 'Mightyena', 263: 'Zigzagoon', 264: 'Linoone', 265: 'Wurmple', 266: 'Silcoon', 267: 'Beautifly', 268: 'Cascoon', 269: 'Dustox', 270: 'Lotad', 271: 'Lombre', 272: 'Ludicolo', 273: 'Seedot', 274: 'Nuzleaf', 275: 'Shiftry', 276: 'Taillow', 277: 'Swellow', 278: 'Wingull', 279: 'Pelipper', 280: 'Ralts', 281: 'Kirlia', 282: 'Gardevoir', 283: 'Surskit', 284: 'Masquerain', 285: 'Shroomish', 286: 'Breloom', 287: 'Slakoth', 288: 'Vigoroth', 289: 'Slaking', 290: 'Nincada', 291: 'Ninjask', 292: 'Shedinja', 293: 'Whismur', 294: 'Loudred', 295: 'Exploud', 296: 'Makuhita', 297: 'Hariyama', 298: 'Azurill', 299: 'Nosepass', 300: 'Skitty', 301: 'Delcatty', 302: 'Sableye', 303: 'Mawile', 304: 'Aron', 305: 'Lairon', 306: 'Aggron', 307: 'Meditite', 308: 'Medicham', 309: 'Electrike', 310: 'Manectric', 311: 'Plusle', 312: 'Minun', 313: 'Volbeat', 314: 'Illumise', 315: 'Roselia', 316: 'Gulpin', 317: 'Swalot', 318: 'Carvanha', 319: 'Sharpedo', 320: 'Wailmer', 321: 'Wailord', 322: 'Numel', 323: 'Camerupt', 324: 'Torkoal', 325: 'Spoink', 326: 'Grumpig', 327: 'Spinda', 328: 'Trapinch', 329: 'Vibrava', 330: 'Flygon', 331: 'Cacnea', 332: 'Cacturne', 333: 'Swablu', 334: 'Altaria', 335: 'Zangoose', 336: 'Seviper', 337: 'Lunatone', 338: 'Solrock', 339: 'Barboach', 340: 'Whiscash', 341: 'Corphish', 342: 'Crawdaunt', 343: 'Baltoy', 344: 'Claydol', 345: 'Lileep', 346: 'Cradily', 347: 'Anorith', 348: 'Armaldo', 349: 'Feebas', 350: 'Milotic', 351: 'Castform', 352: 'Kecleon', 353: 'Shuppet', 354: 'Banette', 355: 'Duskull', 356: 'Dusclops', 357: 'Tropius', 358: 'Chimecho', 359: 'Absol', 360: 'Wynaut', 361: 'Snorunt', 362: 'Glalie', 363: 'Spheal', 364: 'Sealeo', 365: 'Walrein', 366: 'Clamperl', 367: 'Huntail', 368: 'Gorebyss', 369: 'Relicanth', 370: 'Luvdisc', 371: 'Bagon', 372: 'Shelgon', 373: 'Salamence', 374: 'Beldum', 375: 'Metang', 376: 'Metagross', 377: 'Regirock', 378: 'Regice', 379: 'Registeel', 380: 'Latias', 381: 'Latios', 382: 'Kyogre', 383: 'Groudon', 384: 'Rayquaza', 385: 'Jirachi', 386: 'Deoxys', 387: 'Turtwig', 388: 'Grotle', 389: 'Torterra', 390: 'Chimchar', 391: 'Monferno', 392: 'Infernape', 393: 'Piplup', 394: 'Prinplup', 395: 'Empoleon', 396: 'Starly', 397: 'Staravia', 398: 'Staraptor', 399: 'Bidoof', 400: 'Bibarel', 401: 'Kricketot', 402: 'Kricketune', 403: 'Shinx', 404: 'Luxio', 405: 'Luxray', 406: 'Budew', 407: 'Roserade', 408: 'Cranidos', 409: 'Rampardos', 410: 'Shieldon', 411: 'Bastiodon', 412: 'Burmy', 413: 'Wormadam', 414: 'Mothim', 415: 'Combee', 416: 'Vespiquen', 417: 'Pachirisu', 418: 'Buizel', 419: 'Floatzel', 420: 'Cherubi', 421: 'Cherrim', 422: 'Shellos', 423: 'Gastrodon', 424: 'Ambipom', 425: 'Drifloon', 426: 'Drifblim', 427: 'Buneary', 428: 'Lopunny', 429: 'Mismagius', 430: 'Honchkrow', 431: 'Glameow', 432: 'Purugly', 433: 'Chingling', 434: 'Stunky', 435: 'Skuntank', 436: 'Bronzor', 437: 'Bronzong', 438: 'Bonsly', 439: 'Mime Jr.', 440: 'Happiny', 441: 'Chatot', 442: 'Spiritomb', 443: 'Gible', 444: 'Gabite', 445: 'Garchomp', 446: 'Munchlax', 447: 'Riolu', 448: 'Lucario', 449: 'Hippopotas', 450: 'Hippowdon', 451: 'Skorupi', 452: 'Drapion', 453: 'Croagunk', 454: 'Toxicroak', 455: 'Carnivine', 456: 'Finneon', 457: 'Lumineon', 458: 'Mantyke', 459: 'Snover', 460: 'Abomasnow', 461: 'Weavile', 462: 'Magnezone', 463: 'Lickilicky', 464: 'Rhyperior', 465: 'Tangrowth', 466: 'Electivire', 467: 'Magmortar', 468: 'Togekiss', 469: 'Yanmega', 470: 'Leafeon', 471: 'Glaceon', 472: 'Gliscor', 473: 'Mamoswine', 474: 'Porygon-Z', 475: 'Gallade', 476: 'Probopass', 477: 'Dusknoir', 478: 'Froslass', 479: 'Rotom', 480: 'Uxie', 481: 'Mesprit', 482: 'Azelf', 483: 'Dialga', 484: 'Palkia', 485: 'Heatran', 486: 'Regigigas', 487: 'Giratina', 488: 'Cresselia', 489: 'Phione', 490: 'Manaphy', 491: 'Darkrai', 492: 'Shaymin', 493: 'Arceus', 494: 'Victini', 495: 'Snivy', 496: 'Servine', 497: 'Serperior', 498: 'Tepig', 499: 'Pignite', 500: 'Emboar', 501: 'Oshawott', 502: 'Dewott', 503: 'Samurott', 504: 'Patrat', 505: 'Watchog', 506: 'Lillipup', 507: 'Herdier', 508: 'Stoutland', 509: 'Purrloin', 510: 'Liepard', 511: 'Pansage', 512: 'Simisage', 513: 'Pansear', 514: 'Simisear', 515: 'Panpour', 516: 'Simipour', 517: 'Munna', 518: 'Musharna', 519: 'Pidove', 520: 'Tranquill', 521: 'Unfezant', 522: 'Blitzle', 523: 'Zebstrika', 524: 'Roggenrola', 525: 'Boldore', 526: 'Gigalith', 527: 'Woobat', 528: 'Swoobat', 529: 'Drilbur', 530: 'Excadrill', 531: 'Audino', 532: 'Timburr', 533: 'Gurdurr', 534: 'Conkeldurr', 535: 'Tympole', 536: 'Palpitoad', 537: 'Seismitoad', 538: 'Throh', 539: 'Sawk', 540: 'Sewaddle', 541: 'Swadloon', 542: 'Leavanny', 543: 'Venipede', 544: 'Whirlipede', 545: 'Scolipede', 546: 'Cottonee', 547: 'Whimsicott', 548: 'Petilil', 549: 'Lilligant', 550: 'Basculin', 551: 'Sandile', 552: 'Krokorok', 553: 'Krookodile', 554: 'Darumaka', 555: 'Darmanitan', 556: 'Maractus', 557: 'Dwebble', 558: 'Crustle', 559: 'Scraggy', 560: 'Scrafty', 561: 'Sigilyph', 562: 'Yamask', 563: 'Cofagrigus', 564: 'Tirtouga', 565: 'Carracosta', 566: 'Archen', 567: 'Archeops', 568: 'Trubbish', 569: 'Garbodor', 570: 'Zorua', 571: 'Zoroark', 572: 'Minccino', 573: 'Cinccino', 574: 'Gothita', 575: 'Gothorita', 576: 'Gothitelle', 577: 'Solosis', 578: 'Duosion', 579: 'Reuniclus', 580: 'Ducklett', 581: 'Swanna', 582: 'Vanillite', 583: 'Vanillish', 584: 'Vanilluxe', 585: 'Deerling', 586: 'Sawsbuck', 587: 'Emolga', 588: 'Karrablast', 589: 'Escavalier', 590: 'Foongus', 591: 'Amoonguss', 592: 'Frillish', 593: 'Jellicent', 594: 'Alomomola', 595: 'Joltik', 596: 'Galvantula', 597: 'Ferroseed', 598: 'Ferrothorn', 599: 'Klink', 600: 'Klang', 601: 'Klinklang', 602: 'Tynamo', 603: 'Eelektrik', 604: 'Eelektross', 605: 'Elgyem', 606: 'Beheeyem', 607: 'Litwick', 608: 'Lampent', 609: 'Chandelure', 610: 'Axew', 611: 'Fraxure', 612: 'Haxorus', 613: 'Cubchoo', 614: 'Beartic', 615: 'Cryogonal', 616: 'Shelmet', 617: 'Accelgor', 618: 'Stunfisk', 619: 'Mienfoo', 620: 'Mienshao', 621: 'Druddigon', 622: 'Golett', 623: 'Golurk', 624: 'Pawniard', 625: 'Bisharp', 626: 'Bouffalant', 627: 'Rufflet', 628: 'Braviary', 629: 'Vullaby', 630: 'Mandibuzz', 631: 'Heatmor', 632: 'Durant', 633: 'Deino', 634: 'Zweilous', 635: 'Hydreigon', 636: 'Larvesta', 637: 'Volcarona', 638: 'Cobalion', 639: 'Terrakion', 640: 'Virizion', 641: 'Tornadus', 642: 'Thundurus', 643: 'Reshiram', 644: 'Zekrom', 645: 'Landorus', 646: 'Kyurem', 647: 'Keldeo', 648: 'Meloetta', 649: 'Genesect', 650: 'Chespin', 651: 'Quilladin', 652: 'Chesnaught', 653: 'Fennekin', 654: 'Braixen', 655: 'Delphox', 656: 'Froakie', 657: 'Frogadier', 658: 'Greninja', 659: 'Bunnelby', 660: 'Diggersby', 661: 'Fletchling', 662: 'Fletchinder', 663: 'Talonflame', 664: 'Scatterbug', 665: 'Spewpa', 666: 'Vivillon', 667: 'Litleo', 668: 'Pyroar', 669: 'Flabébé', 670: 'Floette', 671: 'Florges', 672: 'Skiddo', 673: 'Gogoat', 674: 'Pancham', 675: 'Pangoro', 676: 'Furfrou', 677: 'Espurr', 678: 'Meowstic', 679: 'Honedge', 680: 'Doublade', 681: 'Aegislash', 682: 'Spritzee', 683: 'Aromatisse', 684: 'Swirlix', 685: 'Slurpuff', 686: 'Inkay', 687: 'Malamar', 688: 'Binacle', 689: 'Barbaracle', 690: 'Skrelp', 691: 'Dragalge', 692: 'Clauncher', 693: 'Clawitzer', 694: 'Helioptile', 695: 'Heliolisk', 696: 'Tyrunt', 697: 'Tyrantrum', 698: 'Amaura', 699: 'Aurorus', 700: 'Sylveon', 701: 'Hawlucha', 702: 'Dedenne', 703: 'Carbink', 704: 'Goomy', 705: 'Sliggoo', 706: 'Goodra', 707: 'Klefki', 708: 'Phantump', 709: 'Trevenant', 710: 'Pumpkaboo', 711: 'Gourgeist', 712: 'Bergmite', 713: 'Avalugg', 714: 'Noibat', 715: 'Noivern', 716: 'Xerneas', 717: 'Yveltal', 718: 'Zygarde', 719: 'Diancie', 720: 'Hoopa', 721: 'Volcanion'}\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "p_ids = np.arange(1,151+1)\n",
    "names = np.array([pokemon_name_map[p_id] for p_id in p_ids])\n",
    "total_counts = [data.counts[p_id-1].sum() for p_id in p_ids]\n",
    "\n",
    "region_dependent_labels = [region_dependent_label(log_bayes_factors[p_id-1]) for p_id in p_ids]\n",
    "\n",
    "rate_ratio = np.empty(N_POKEMON)\n",
    "alphas = prior_overall.alpha.reshape((N_POKEMON, n_regions))\n",
    "for p_id in p_ids:\n",
    "    mask = p_id != p_ids\n",
    "    rate_ratio[p_id-1] =  (  (data.counts[p_id-1][0] + alphas[p_id-1][0]) \n",
    "                           / (data.counts[p_id-1][1] + alphas[p_id-1][1]) ) \\\n",
    "                        / (  (data.counts[mask].sum(axis=0)[0] + alphas[mask].sum(axis=0)[0]) \n",
    "                           / (data.counts[mask].sum(axis=0)[1] + alphas[mask].sum(axis=0)[1]) )\n",
    "\n",
    "more_likely_in = [db_labels[0] if rate>1 else db_labels[1] for rate in rate_ratio]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "t = Table([p_ids, names, total_counts, log_bayes_factors,\n",
    "           region_dependent_labels, more_likely_in, rate_ratio],\n",
    "          names=(\"Pokemon #\", \"names\", \"total_counts\", \n",
    "                 \"significance proxy\", \"Region Dependent?\",\n",
    "                 \"More Common In\", \n",
    "                 \"spawn rate ratio ({0} / {1})\".format(db_labels[0], db_labels[1])))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Final Results\n",
    "\n",
    "Most columns are self-explanatory, except:\n",
    " - `significance proxy`: values > 0 suggest the spawn rate is different between regions. Large values means we're very confident of that.  Technically this is the log of the Bayes Factor.\n",
    "\n",
    " - Spawn rate ratio is an estimate of the relative spawn rate. It is *approximately, but not exactly*: $$ \\frac{\\text{number of this Pokemon in region 1 } / \\text{ total number of Pokemon in region 1}}{\\text{number of this Pokemon in region 2 } / \\text{ total number of Pokemon in region 2}} $$\n",
    " \n",
    " \n",
    " \n",
    " \n",
    "So if you want to look for a pokemon that's changed a lot, look for a `spawn rate ratio` far from 1 **and** a large value for `significance proxy`. One easy way to do this is sort by `spawn rate ratio` and then sort by `Region Dependent?`.\n",
    "\n",
    "In my example notebook you'll know you're doing this right when either Clefairy or Magmar comes to the top fo the list."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<script src=\"https://ajax.googleapis.com/ajax/libs/jquery/1.12.4/jquery.min.js\"></script>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%HTML # needed to make the table sortable in nbviewer\n",
    "<script src=\"https://ajax.googleapis.com/ajax/libs/jquery/1.12.4/jquery.min.js\"></script>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "&lt;Table length=151&gt;\n",
       "<table id=\"table4704613584-30734\" class=\"table-striped table-bordered table-condensed\">\n",
       "<thead><tr><th>Pokemon #</th><th>names</th><th>total_counts</th><th>significance proxy</th><th>Region Dependent?</th><th>More Common In</th><th>spawn rate ratio (Paris / Santa Cruz)</th></tr></thead>\n",
       "<tr><td>1</td><td>Bulbasaur</td><td>44</td><td>6.77094507132</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.274552536912</td></tr>\n",
       "<tr><td>2</td><td>Ivysaur</td><td>2</td><td>0.900364678959</td><td>Not sure</td><td>Santa Cruz</td><td>0.128284115036</td></tr>\n",
       "<tr><td>3</td><td>Venusaur</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>4</td><td>Charmander</td><td>22</td><td>11.3079079351</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.0781433170468</td></tr>\n",
       "<tr><td>5</td><td>Charmeleon</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>6</td><td>Charizard</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>7</td><td>Squirtle</td><td>95</td><td>-1.74228009716</td><td>No</td><td>Santa Cruz</td><td>0.773771335815</td></tr>\n",
       "<tr><td>8</td><td>Wartortle</td><td>12</td><td>-1.4067843265</td><td>No</td><td>Paris</td><td>1.21180059466</td></tr>\n",
       "<tr><td>9</td><td>Blastoise</td><td>1</td><td>0.247431359412</td><td>Not sure</td><td>Santa Cruz</td><td>0.213819725922</td></tr>\n",
       "<tr><td>10</td><td>Caterpie</td><td>915</td><td>136.15121324</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.320822653707</td></tr>\n",
       "<tr><td>11</td><td>Metapod</td><td>52</td><td>11.6853135476</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.21884136051</td></tr>\n",
       "<tr><td>12</td><td>Butterfree</td><td>1</td><td>0.247431359412</td><td>Not sure</td><td>Santa Cruz</td><td>0.213819725922</td></tr>\n",
       "<tr><td>13</td><td>Weedle</td><td>1816</td><td>231.290508095</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.34893039532</td></tr>\n",
       "<tr><td>14</td><td>Kakuna</td><td>126</td><td>8.98703323009</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.422350893558</td></tr>\n",
       "<tr><td>15</td><td>Beedrill</td><td>14</td><td>2.9896631039</td><td>Yes</td><td>Santa Cruz</td><td>0.195131817922</td></tr>\n",
       "<tr><td>16</td><td>Pidgey</td><td>10241</td><td>59.4541140904</td><td>Yes - very confident</td><td>Paris</td><td>1.3079487553</td></tr>\n",
       "<tr><td>17</td><td>Pidgeotto</td><td>658</td><td>-3.45133944105</td><td>No</td><td>Paris</td><td>1.01330765508</td></tr>\n",
       "<tr><td>18</td><td>Pidgeot</td><td>81</td><td>2.01710674339</td><td>Yes</td><td>Paris</td><td>2.0586589435</td></tr>\n",
       "<tr><td>19</td><td>Rattata</td><td>8930</td><td>489.949207996</td><td>Yes - very confident</td><td>Paris</td><td>2.2538027074</td></tr>\n",
       "<tr><td>20</td><td>Raticate</td><td>269</td><td>12.622710503</td><td>Yes - very confident</td><td>Paris</td><td>2.13891820375</td></tr>\n",
       "<tr><td>21</td><td>Spearow</td><td>4331</td><td>65.9878212252</td><td>Yes - very confident</td><td>Paris</td><td>1.49683135</td></tr>\n",
       "<tr><td>22</td><td>Fearow</td><td>125</td><td>-0.205359889659</td><td>Not sure</td><td>Paris</td><td>1.51571290235</td></tr>\n",
       "<tr><td>23</td><td>Ekans</td><td>302</td><td>65.4992892633</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.245626974018</td></tr>\n",
       "<tr><td>24</td><td>Arbok</td><td>11</td><td>1.11332471245</td><td>Yes</td><td>Santa Cruz</td><td>0.264049562378</td></tr>\n",
       "<tr><td>25</td><td>Pikachu</td><td>21</td><td>1.50667846199</td><td>Yes</td><td>Santa Cruz</td><td>0.331619270294</td></tr>\n",
       "<tr><td>26</td><td>Raichu</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>27</td><td>Sandshrew</td><td>18</td><td>-1.68503733585</td><td>No</td><td>Santa Cruz</td><td>0.983633274118</td></tr>\n",
       "<tr><td>28</td><td>Sandslash</td><td>1</td><td>0.247431359412</td><td>Not sure</td><td>Santa Cruz</td><td>0.213819725922</td></tr>\n",
       "<tr><td>29</td><td>Nidoran(F)</td><td>116</td><td>33.2910369348</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.187012272798</td></tr>\n",
       "<tr><td>30</td><td>Nidorina</td><td>4</td><td>2.46585308196</td><td>Yes</td><td>Santa Cruz</td><td>0.0712603744451</td></tr>\n",
       "<tr><td>31</td><td>Nidoqueen</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>32</td><td>Nidoran(M)</td><td>118</td><td>72.7472381739</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.0553092242521</td></tr>\n",
       "<tr><td>33</td><td>Nidorino</td><td>15</td><td>12.1778271318</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.0206747981147</td></tr>\n",
       "<tr><td>34</td><td>Nidoking</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>35</td><td>Clefairy</td><td>35</td><td>30.5885858391</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.00901615056096</td></tr>\n",
       "<tr><td>36</td><td>Clefable</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>37</td><td>Vulpix</td><td>8</td><td>-0.402610295119</td><td>Not sure</td><td>Santa Cruz</td><td>0.408150300784</td></tr>\n",
       "<tr><td>38</td><td>Ninetales</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>39</td><td>Jigglypuff</td><td>246</td><td>7.22480986532</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.561495819555</td></tr>\n",
       "<tr><td>40</td><td>Wigglytuff</td><td>8</td><td>-1.09963482369</td><td>No</td><td>Santa Cruz</td><td>0.641442415351</td></tr>\n",
       "<tr><td>41</td><td>Zubat</td><td>6179</td><td>44.3234137813</td><td>Yes - very confident</td><td>Paris</td><td>1.33035404975</td></tr>\n",
       "<tr><td>42</td><td>Golbat</td><td>275</td><td>4.89008939915</td><td>Yes</td><td>Paris</td><td>1.67816623265</td></tr>\n",
       "<tr><td>43</td><td>Oddish</td><td>100</td><td>32.6558859916</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.162701965334</td></tr>\n",
       "<tr><td>44</td><td>Gloom</td><td>5</td><td>0.658167702051</td><td>Not sure</td><td>Santa Cruz</td><td>0.213789376158</td></tr>\n",
       "<tr><td>45</td><td>Vileplume</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>46</td><td>Paras</td><td>464</td><td>126.646070121</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.203654016021</td></tr>\n",
       "<tr><td>47</td><td>Parasect</td><td>16</td><td>4.35896326762</td><td>Yes</td><td>Santa Cruz</td><td>0.166203381978</td></tr>\n",
       "<tr><td>48</td><td>Venonat</td><td>498</td><td>68.3168449623</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.334495203988</td></tr>\n",
       "<tr><td>49</td><td>Venomoth</td><td>13</td><td>0.891927495123</td><td>Not sure</td><td>Santa Cruz</td><td>0.303749702816</td></tr>\n",
       "<tr><td>50</td><td>Diglett</td><td>55</td><td>-1.76190041548</td><td>No</td><td>Paris</td><td>1.30078276823</td></tr>\n",
       "<tr><td>51</td><td>Dugtrio</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>52</td><td>Meowth</td><td>213</td><td>-0.759509526724</td><td>Not sure</td><td>Santa Cruz</td><td>0.751111574746</td></tr>\n",
       "<tr><td>53</td><td>Persian</td><td>1</td><td>-0.198153063264</td><td>Not sure</td><td>Paris</td><td>1.92456763434</td></tr>\n",
       "<tr><td>54</td><td>Psyduck</td><td>524</td><td>152.738644547</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.191416993671</td></tr>\n",
       "<tr><td>55</td><td>Golduck</td><td>13</td><td>6.70075780547</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.0769271155034</td></tr>\n",
       "<tr><td>56</td><td>Mankey</td><td>32</td><td>-0.0352089437278</td><td>Not sure</td><td>Santa Cruz</td><td>0.502522529482</td></tr>\n",
       "<tr><td>57</td><td>Primeape</td><td>1</td><td>0.247431359412</td><td>Not sure</td><td>Santa Cruz</td><td>0.213819725922</td></tr>\n",
       "<tr><td>58</td><td>Growlithe</td><td>51</td><td>16.0955914319</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.161968747746</td></tr>\n",
       "<tr><td>59</td><td>Arcanine</td><td>1</td><td>-0.198153063264</td><td>Not sure</td><td>Paris</td><td>1.92456763434</td></tr>\n",
       "<tr><td>60</td><td>Poliwag</td><td>479</td><td>4.35421118601</td><td>Yes</td><td>Santa Cruz</td><td>0.69727721921</td></tr>\n",
       "<tr><td>61</td><td>Poliwhirl</td><td>23</td><td>0.42466497275</td><td>Not sure</td><td>Santa Cruz</td><td>0.420083517092</td></tr>\n",
       "<tr><td>62</td><td>Poliwrath</td><td>1</td><td>0.247431359412</td><td>Not sure</td><td>Santa Cruz</td><td>0.213819725922</td></tr>\n",
       "<tr><td>63</td><td>Abra</td><td>66</td><td>3.07997238044</td><td>Yes</td><td>Santa Cruz</td><td>0.446029127914</td></tr>\n",
       "<tr><td>64</td><td>Kadabra</td><td>2</td><td>0.00917409758655</td><td>Not sure</td><td>Paris</td><td>3.20773655561</td></tr>\n",
       "<tr><td>65</td><td>Alakazam</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>66</td><td>Machop</td><td>33</td><td>-1.90269631683</td><td>No</td><td>Santa Cruz</td><td>0.8626097984</td></tr>\n",
       "<tr><td>67</td><td>Machoke</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>68</td><td>Machamp</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>69</td><td>Bellsprout</td><td>176</td><td>91.8798387125</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.0833055521442</td></tr>\n",
       "<tr><td>70</td><td>Weepinbell</td><td>3</td><td>1.65869536916</td><td>Yes</td><td>Santa Cruz</td><td>0.0916259960846</td></tr>\n",
       "<tr><td>71</td><td>Victreebel</td><td>1</td><td>0.247431359412</td><td>Not sure</td><td>Santa Cruz</td><td>0.213819725922</td></tr>\n",
       "<tr><td>72</td><td>Tentacool</td><td>86</td><td>15.4912559096</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.265266520903</td></tr>\n",
       "<tr><td>73</td><td>Tentacruel</td><td>7</td><td>2.01076516479</td><td>Yes</td><td>Santa Cruz</td><td>0.147990212382</td></tr>\n",
       "<tr><td>74</td><td>Geodude</td><td>8</td><td>2.7439018948</td><td>Yes</td><td>Santa Cruz</td><td>0.128250463249</td></tr>\n",
       "<tr><td>75</td><td>Graveler</td><td>1</td><td>-0.198153063264</td><td>Not sure</td><td>Paris</td><td>1.92456763434</td></tr>\n",
       "<tr><td>76</td><td>Golem</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>77</td><td>Ponyta</td><td>4</td><td>-0.817056912903</td><td>Not sure</td><td>Paris</td><td>1.49691143541</td></tr>\n",
       "<tr><td>78</td><td>Rapidash</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>79</td><td>Slowpoke</td><td>237</td><td>70.7970366081</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.184334466739</td></tr>\n",
       "<tr><td>80</td><td>Slowbro</td><td>5</td><td>0.658167702051</td><td>Not sure</td><td>Santa Cruz</td><td>0.213789376158</td></tr>\n",
       "<tr><td>81</td><td>Magnemite</td><td>5</td><td>1.07307903765</td><td>Yes</td><td>Paris</td><td>7.05783783784</td></tr>\n",
       "<tr><td>82</td><td>Magneton</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>83</td><td>Farfetch&apos;d</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>84</td><td>Doduo</td><td>60</td><td>22.6999221924</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.133030765907</td></tr>\n",
       "<tr><td>85</td><td>Dodrio</td><td>2</td><td>0.900364678959</td><td>Not sure</td><td>Santa Cruz</td><td>0.128284115036</td></tr>\n",
       "<tr><td>86</td><td>Seel</td><td>9</td><td>2.77063230772</td><td>Yes</td><td>Paris</td><td>12.1926938523</td></tr>\n",
       "<tr><td>87</td><td>Dewgong</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>88</td><td>Grimer</td><td>8</td><td>-1.293931312</td><td>No</td><td>Paris</td><td>1.00808052951</td></tr>\n",
       "<tr><td>89</td><td>Muk</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>90</td><td>Shellder</td><td>9</td><td>2.77063230772</td><td>Yes</td><td>Paris</td><td>12.1926938523</td></tr>\n",
       "<tr><td>91</td><td>Cloyster</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>92</td><td>Gastly</td><td>125</td><td>-2.34616781019</td><td>No</td><td>Santa Cruz</td><td>0.868962071624</td></tr>\n",
       "<tr><td>93</td><td>Haunter</td><td>5</td><td>-1.0804851292</td><td>No</td><td>Santa Cruz</td><td>0.898092811366</td></tr>\n",
       "<tr><td>94</td><td>Gengar</td><td>1</td><td>-0.198153063264</td><td>Not sure</td><td>Paris</td><td>1.92456763434</td></tr>\n",
       "<tr><td>95</td><td>Onix</td><td>1</td><td>-0.198153063264</td><td>Not sure</td><td>Paris</td><td>1.92456763434</td></tr>\n",
       "<tr><td>96</td><td>Drowzee</td><td>37</td><td>-1.33909471919</td><td>No</td><td>Paris</td><td>1.47873968653</td></tr>\n",
       "<tr><td>97</td><td>Hypno</td><td>2</td><td>0.00917409758655</td><td>Not sure</td><td>Paris</td><td>3.20773655561</td></tr>\n",
       "<tr><td>98</td><td>Krabby</td><td>422</td><td>2.20773955613</td><td>Yes</td><td>Santa Cruz</td><td>0.723131127238</td></tr>\n",
       "<tr><td>99</td><td>Kingler</td><td>13</td><td>-0.945943537631</td><td>Not sure</td><td>Santa Cruz</td><td>0.555859299587</td></tr>\n",
       "<tr><td>100</td><td>Voltorb</td><td>12</td><td>1.42535958863</td><td>Yes</td><td>Paris</td><td>4.9199428439</td></tr>\n",
       "<tr><td>101</td><td>Electrode</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>102</td><td>Exeggcute</td><td>96</td><td>14.2151141514</td><td>Yes - very confident</td><td>Paris</td><td>4.34749080206</td></tr>\n",
       "<tr><td>103</td><td>Exeggutor</td><td>1</td><td>0.247431359412</td><td>Not sure</td><td>Santa Cruz</td><td>0.213819725922</td></tr>\n",
       "<tr><td>104</td><td>Cubone</td><td>23</td><td>-1.43885488636</td><td>No</td><td>Santa Cruz</td><td>0.697141580988</td></tr>\n",
       "<tr><td>105</td><td>Marowak</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>106</td><td>Hitmonlee</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>107</td><td>Hitmonchan</td><td>1</td><td>-0.198153063264</td><td>Not sure</td><td>Paris</td><td>1.92456763434</td></tr>\n",
       "<tr><td>108</td><td>Lickitung</td><td>241</td><td>49.1348691082</td><td>Yes - very confident</td><td>Paris</td><td>5.73474313425</td></tr>\n",
       "<tr><td>109</td><td>Koffing</td><td>279</td><td>-2.67253999549</td><td>No</td><td>Paris</td><td>1.11166348137</td></tr>\n",
       "<tr><td>110</td><td>Weezing</td><td>11</td><td>-0.336761047482</td><td>Not sure</td><td>Paris</td><td>2.43824528885</td></tr>\n",
       "<tr><td>111</td><td>Rhyhorn</td><td>15</td><td>0.735823592323</td><td>Not sure</td><td>Santa Cruz</td><td>0.335885273028</td></tr>\n",
       "<tr><td>112</td><td>Rhydon</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>113</td><td>Chansey</td><td>2</td><td>0.900364678959</td><td>Not sure</td><td>Santa Cruz</td><td>0.128284115036</td></tr>\n",
       "<tr><td>114</td><td>Tangela</td><td>6</td><td>-0.471367078772</td><td>Not sure</td><td>Paris</td><td>2.35247104247</td></tr>\n",
       "<tr><td>115</td><td>Kangaskhan</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>116</td><td>Horsea</td><td>213</td><td>3.05801214629</td><td>Yes</td><td>Santa Cruz</td><td>0.622228771479</td></tr>\n",
       "<tr><td>117</td><td>Seadra</td><td>4</td><td>-0.882315095613</td><td>Not sure</td><td>Santa Cruz</td><td>0.641470099988</td></tr>\n",
       "<tr><td>118</td><td>Goldeen</td><td>404</td><td>1.81559523891</td><td>Yes</td><td>Santa Cruz</td><td>0.727252270064</td></tr>\n",
       "<tr><td>119</td><td>Seaking</td><td>12</td><td>1.70928685305</td><td>Yes</td><td>Santa Cruz</td><td>0.236240647948</td></tr>\n",
       "<tr><td>120</td><td>Staryu</td><td>419</td><td>1.0653069731</td><td>Yes</td><td>Santa Cruz</td><td>0.748711507095</td></tr>\n",
       "<tr><td>121</td><td>Starmie</td><td>7</td><td>-0.76790130771</td><td>Not sure</td><td>Santa Cruz</td><td>0.498880395336</td></tr>\n",
       "<tr><td>122</td><td>Mr. Mime</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>123</td><td>Scyther</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>124</td><td>Jynx</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>125</td><td>Electabuzz</td><td>5</td><td>-0.678715990719</td><td>Not sure</td><td>Paris</td><td>1.92467472298</td></tr>\n",
       "<tr><td>126</td><td>Magmar</td><td>21</td><td>5.02871183626</td><td>Yes - very confident</td><td>Paris</td><td>8.77338226772</td></tr>\n",
       "<tr><td>127</td><td>Pinsir</td><td>61</td><td>-2.23655732589</td><td>No</td><td>Santa Cruz</td><td>0.918116846646</td></tr>\n",
       "<tr><td>128</td><td>Tauros</td><td>2</td><td>0.900364678959</td><td>Not sure</td><td>Santa Cruz</td><td>0.128284115036</td></tr>\n",
       "<tr><td>129</td><td>Magikarp</td><td>973</td><td>254.748094013</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.212172447165</td></tr>\n",
       "<tr><td>130</td><td>Gyarados</td><td>2</td><td>0.900364678959</td><td>Not sure</td><td>Santa Cruz</td><td>0.128284115036</td></tr>\n",
       "<tr><td>131</td><td>Lapras</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>132</td><td>Ditto</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>133</td><td>Eevee</td><td>568</td><td>145.610862392</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.215995484611</td></tr>\n",
       "<tr><td>134</td><td>Vaporeon</td><td>6</td><td>1.31118799368</td><td>Yes</td><td>Santa Cruz</td><td>0.174908052108</td></tr>\n",
       "<tr><td>135</td><td>Jolteon</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>136</td><td>Flareon</td><td>1</td><td>-0.198153063264</td><td>Not sure</td><td>Paris</td><td>1.92456763434</td></tr>\n",
       "<tr><td>137</td><td>Porygon</td><td>10</td><td>3.21446920821</td><td>Yes</td><td>Paris</td><td>13.476655725</td></tr>\n",
       "<tr><td>138</td><td>Omanyte</td><td>4</td><td>-0.817056912903</td><td>Not sure</td><td>Paris</td><td>1.49691143541</td></tr>\n",
       "<tr><td>139</td><td>Omastar</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>140</td><td>Kabuto</td><td>4</td><td>-0.882315095613</td><td>Not sure</td><td>Santa Cruz</td><td>0.641470099988</td></tr>\n",
       "<tr><td>141</td><td>Kabutops</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>142</td><td>Aerodactyl</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>143</td><td>Snorlax</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>144</td><td>Articuno</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>145</td><td>Zapdos</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>146</td><td>Moltres</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>147</td><td>Dratini</td><td>59</td><td>13.1119627757</td><td>Yes - very confident</td><td>Santa Cruz</td><td>0.222980544308</td></tr>\n",
       "<tr><td>148</td><td>Dragonair</td><td>1</td><td>-0.198153063264</td><td>Not sure</td><td>Paris</td><td>1.92456763434</td></tr>\n",
       "<tr><td>149</td><td>Dragonite</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>150</td><td>Mewtwo</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "<tr><td>151</td><td>Mew</td><td>0</td><td>0.0</td><td>Not sure</td><td>Santa Cruz</td><td>0.641497780351</td></tr>\n",
       "</table><style>table.dataTable {clear: both; width: auto !important; margin: 0 !important;}\n",
       ".dataTables_info, .dataTables_length, .dataTables_filter, .dataTables_paginate{\n",
       "display: inline-block; margin-right: 1em; }\n",
       ".paginate_button { margin-right: 5px; }\n",
       "</style>\n",
       "<script>\n",
       "require.config({paths: {\n",
       "    datatables: 'https://cdn.datatables.net/1.10.9/js/jquery.dataTables.min'\n",
       "}});\n",
       "require([\"datatables\"], function(){\n",
       "    console.log(\"$('#table4704613584-30734').dataTable()\");\n",
       "    $('#table4704613584-30734').dataTable({\n",
       "        \"order\": [],\n",
       "        \"iDisplayLength\": 151,\n",
       "        \"aLengthMenu\": [[151, 10, 25, 50, 100, 500, 1000, -1], [151, 10, 25, 50, 100, 500, 1000, 'All']],\n",
       "        \"pagingType\": \"full_numbers\"\n",
       "    });\n",
       "});\n",
       "</script>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t.show_in_notebook(show_row_index=False, display_length=N_POKEMON)"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [poke27]",
   "language": "python",
   "name": "Python [poke27]"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}

db_Paris.sqlite

db_SC.sqlite