2-topic-modelling-pokemon.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "f70924e8-c977-4a85-bfd3-b4c950371f4f",
   "metadata": {
    "tags": []
   },
   "source": [
    "# Pok&eacute;Map\n",
    "Built using the Tutte Institute Data Science stack: https://github.com/tutteinstitute. <br>\n",
    "Data taken from Pok&eacute;API: https://pokeapi.co/.\n",
    " \n",
    "Author: Kaleb D. Ruscitti."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3aef7334-bec4-4827-a920-f76d65f72097",
   "metadata": {},
   "source": [
    "## Topic Modelling\n",
    "\n",
    "Now that you've learned how to visualize data (see tutorial-recipes-bag-of-words.ipynb !), we're going to try sorting our data into topics.\n",
    "\n",
    "In this case, our data points will be Pokemon, fetched gratefully from PokeAPI. A <i> topic </i> in this context means a group of pokemon which are similar in some way. Precisely in what way? Well that will depend on how we model Pokemon as mathematical objects, which we will discuss below.\n",
    "\n",
    "In modern topic modelling, a baseline to define a topic is a <i> cluster </i>, which is a region of relative density in the space of data. Typically, one looks for these regions in the 'low-space', meaning the space of data after it has been dimensionally reduced. Thus, our starting point is to use what we learned in the previous tutorial to find a low-dimensional representation of our data.\n",
    "\n",
    "Once we've run a clustering algorithm, we will have the clusters helpfully labeled with... an integer... In order to extract meaningful information out of the clustering result, we will want to find a better way to <i> represent </i> our clusters. This requires us to look inside the cluster and interpret what we see. \n",
    "\n",
    "All said, when we're doing topic modelling we can add two more steps to the pipeline from before: vectorize, information weight, UMAP, cluster, represent.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "52fa7c1b-6866-40ae-be97-40db4c4f6d92",
   "metadata": {},
   "source": [
    "## Refresher: Vectorize, Information Weight, UMAP\n",
    "\n",
    "In the previous tutorial, you learned that the basic recipe for visualizing data is to turn your data points into vectors, apply the information weight transform, and then use a dimensional reduction algorithm like UMAP.\n",
    "\n",
    "One again we're going to go through these exact steps, but this time we'll talk a little bit more about the choices involved in the vectorizing and dimension reduction steps. Each step has some modelling choices or parameters that need to be selected, and these decisions can affect the final output of the data. Hopefully, you can use your understanding of the problem to find some set of parameters which are more justified that other. Moreover, we can try different sets of choices and try to identify patterns in the data which are invariant under these choices."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "453ff2e0-a5f4-4cdf-9636-b421ae11ebcd",
   "metadata": {
    "tags": []
   },
   "source": [
    "\n",
    "## Imports\n",
    "If any packages are installed you will need to restart the kernel."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "30e4ceb3-ae65-4573-871b-c00210d88d2b",
   "metadata": {
    "jupyter": {
     "source_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/kdruscit/.conda/envs/sklearn/lib/python3.12/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n",
      "/home/kdruscit/.conda/envs/sklearn/lib/python3.12/site-packages/dask/dataframe/__init__.py:31: FutureWarning: \n",
      "Dask dataframe query planning is disabled because dask-expr is not installed.\n",
      "\n",
      "You can install it with `pip install dask[dataframe]` or `conda install dask`.\n",
      "This will raise in a future version.\n",
      "\n",
      "  warnings.warn(msg, FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "try:\n",
    "    from umap import UMAP\n",
    "    from hdbscan import HDBSCAN\n",
    "    from vectorizers import NgramVectorizer\n",
    "    from sklearn.feature_extraction.text import CountVectorizer\n",
    "    from vectorizers.transformers import InformationWeightTransformer\n",
    "    from sklearn.decomposition import PCA\n",
    "    import datamapplot\n",
    "except:\n",
    "    %pip install umap-learn vectorizers hdbscan datamapplot\n",
    "    print(\"Please restart the kernel!\")\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b71533a7-8738-42b8-8c03-501777337186",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Loading Data\n",
    "Load the file `pokedex.feather` included in this repository. The file `fetch.py` can be modified to obtain other data from PokeAPI if you wish."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "acaad41b-12f3-4290-bca7-b18e96f599c3",
   "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>name</th>\n",
       "      <th>type</th>\n",
       "      <th>abilities</th>\n",
       "      <th>moves</th>\n",
       "      <th>stats</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>bulbasaur</td>\n",
       "      <td>[grass, poison]</td>\n",
       "      <td>[overgrow, chlorophyll]</td>\n",
       "      <td>[razor-wind, swords-dance, cut, bind, vine-whi...</td>\n",
       "      <td>[45, 49, 49, 65, 65, 45]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ivysaur</td>\n",
       "      <td>[grass, poison]</td>\n",
       "      <td>[overgrow, chlorophyll]</td>\n",
       "      <td>[swords-dance, cut, bind, vine-whip, headbutt,...</td>\n",
       "      <td>[60, 62, 63, 80, 80, 60]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>venusaur</td>\n",
       "      <td>[grass, poison]</td>\n",
       "      <td>[overgrow, chlorophyll]</td>\n",
       "      <td>[swords-dance, cut, bind, vine-whip, headbutt,...</td>\n",
       "      <td>[80, 82, 83, 100, 100, 80]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>charmander</td>\n",
       "      <td>[fire]</td>\n",
       "      <td>[blaze, solar-power]</td>\n",
       "      <td>[mega-punch, fire-punch, thunder-punch, scratc...</td>\n",
       "      <td>[39, 52, 43, 60, 50, 65]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>charmeleon</td>\n",
       "      <td>[fire]</td>\n",
       "      <td>[blaze, solar-power]</td>\n",
       "      <td>[mega-punch, fire-punch, thunder-punch, scratc...</td>\n",
       "      <td>[58, 64, 58, 80, 65, 80]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         name             type                abilities  \\\n",
       "0   bulbasaur  [grass, poison]  [overgrow, chlorophyll]   \n",
       "1     ivysaur  [grass, poison]  [overgrow, chlorophyll]   \n",
       "2    venusaur  [grass, poison]  [overgrow, chlorophyll]   \n",
       "3  charmander           [fire]     [blaze, solar-power]   \n",
       "4  charmeleon           [fire]     [blaze, solar-power]   \n",
       "\n",
       "                                               moves  \\\n",
       "0  [razor-wind, swords-dance, cut, bind, vine-whi...   \n",
       "1  [swords-dance, cut, bind, vine-whip, headbutt,...   \n",
       "2  [swords-dance, cut, bind, vine-whip, headbutt,...   \n",
       "3  [mega-punch, fire-punch, thunder-punch, scratc...   \n",
       "4  [mega-punch, fire-punch, thunder-punch, scratc...   \n",
       "\n",
       "                        stats  \n",
       "0    [45, 49, 49, 65, 65, 45]  \n",
       "1    [60, 62, 63, 80, 80, 60]  \n",
       "2  [80, 82, 83, 100, 100, 80]  \n",
       "3    [39, 52, 43, 60, 50, 65]  \n",
       "4    [58, 64, 58, 80, 65, 80]  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pokedex = pd.read_feather('data/pokedex.feather')\n",
    "pokedex.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b68f3e7d-b5a8-4253-9578-3cbe894476b1",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Vectorizing\n",
    "In order to do any data science with pokemon, we need a way to turn a pokemon into a vector. There are many attributes associated to each Pokemon. They have heights and weights, colours, types, moves, stats, pokedex entries and so on. The first modelling choice that the data scientist must make is\n",
    "\n",
    "` Which attributes of my data are relevant to the question I am studying? `\n",
    "\n",
    "In this case, I want to group Pokemon by their use in Pokemon battles. Many attributes of the Pokemon, such as their height, egg group, pokedex entries, have no affect on the gameplay of a Pokemon battle. Some of the key attributes that do affect battles are the Pokemon's moves, type, abilities and stats, and so those are the attributes that I have included in my dataset.\n",
    "\n",
    "One could instead use attributes like weight, height, colour and egg group to include information about the Pokemon's appearance. In this case, our groups of Pokemon would be ones which look similar. When communicating the results of your data analysis to others, it is important to make sure that you specify the modelling choices that you made, and explain how they affect your results!\n",
    "\n",
    "Once I have my dataframe of attributes that I wish to include, we must turn these into vectors. Once again, we will use the <i> bag of words</i> approach. \n",
    "\n",
    "The CountVectorizer object from scikit-learn will take in a list of words and turn it into a vector that counts the number of appearances of each word. Therefore we just need to turn each pokemon into a list of words. For moves, abilities and type this is straightforward:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "0d00ad9b-710b-41a8-8190-66c1c39f2768",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Pikachu's tokens:\n",
      "static lightning_rod  mega_punch pay_day thunder_punch slam double_kick mega_kick headbutt body_slam take_down double_edge tail_whip growl surf submission counter seismic_toss strength thunder_shock thunderbolt thunder_wave thunder dig toxic agility quick_attack rage mimic double_team defense_curl light_screen reflect bide swift skull_bash flash rest substitute thief snore curse reversal protect sweet_kiss mud_slap zap_cannon detect endure charm rollout swagger spark attract sleep_talk return frustration dynamic_punch encore iron_tail hidden_power rain_dance rock_smash uproar facade focus_punch charge helping_hand brick_break knock_off endeavor secret_power fake_tears signal_beam covet volt_tackle calm_mind shock_wave natural_gift feint fling magnet_rise nasty_plot discharge captivate grass_knot charge_beam electro_ball round echoed_voice volt_switch electroweb wild_charge disarming_voice draining_kiss play_rough play_nice confide eerie_impulse electric_terrain nuzzle laser_focus rising_voltage tera_blast trailblaze alluring_voice upper_hand \n"
     ]
    }
   ],
   "source": [
    "## Appending names, moves & abilities together.\n",
    "tokens = []\n",
    "for i, name in enumerate(pokedex['name'].to_list()):\n",
    "    abil_tokens = ''.join(x.replace(\"-\",\"_\")+' ' for x in list(pokedex['abilities'].iloc[i]))\n",
    "    move_tokens = ''.join(x.replace(\"-\",\"_\")+' ' for x in list(pokedex['moves'].iloc[i]))\n",
    "    tokens.append(abil_tokens + ' ' + move_tokens)\n",
    "\n",
    "\n",
    "pokedex['tokens']=tokens\n",
    "print(\"Pikachu's tokens:\")\n",
    "print(pokedex['tokens'].iloc[24])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ff9fd740-0ccc-4397-888c-08f40db0feb2",
   "metadata": {},
   "source": [
    "However a pokemon's stats are not categorical data like moves and abilities. Therefore we need to make a choice on how to include a pokemon's stats in the data. I've decided to turn the stats into additional words by taking the top two stats of any given pokemon.\n",
    "\n",
    "We can throw these extra words into the CountVectorizer, however we can do slightly better. If we count digrams (pairs of words) instead of single word count, we can keep track of the order of the top two stats, and therefore get a more refined embedding of the pokemon as vectors. This means that DEF ATK and ATK DEF will be considered different objects. However we also want to remember that DEF ATK and ATK DEF are closer together then DEF ATK and SPA SPE; so we will also count the individual words. To do this we pass the parameter `ngram_range=(1,2)` to the second CountVectorizer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "94f6fae5-ae7a-4cf7-abfa-62d8b13ec371",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Turning stats into digrams\n",
    "stat_arr = pokedex['stats'].to_numpy()\n",
    "stat_names = ['HP','ATK','DEF','SPA','SPD','SPE']\n",
    "top_stats = []\n",
    "second_stats = []\n",
    "for row in stat_arr:\n",
    "    if np.amax(row) == np.amin(row):\n",
    "        top_stats.append('ALL')\n",
    "        second_stats.append('')\n",
    "    else:\n",
    "        srt=np.argsort(row)\n",
    "        top_stats.append(stat_names[srt[-1]])\n",
    "        second_stats.append(stat_names[srt[-2]])\n",
    "    \n",
    "pokedex['top_stat'] = top_stats\n",
    "pokedex['second_stat'] = second_stats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "a382c5e1-fc18-4c0f-8320-5c9487cce8aa",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1025, 1080)\n"
     ]
    }
   ],
   "source": [
    "vectorizer = CountVectorizer()\n",
    "token_vectors = vectorizer.fit_transform(pokedex['tokens'].to_list())\n",
    "# vectors is a sparse matrix, but we don't have that much data so let's densify it.\n",
    "token_vectors = token_vectors.toarray()\n",
    "print(np.shape(token_vectors))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "f3532bd1-a549-48b7-b43e-3d0aac31301f",
   "metadata": {},
   "outputs": [],
   "source": [
    "stat_vectorizer = CountVectorizer(ngram_range=(1,2))\n",
    "tmp = (pokedex['top_stat'] +\" \"+ pokedex['second_stat']).to_list()\n",
    "stat_vectors = stat_vectorizer.fit_transform(tmp)\n",
    "stat_vectors = stat_vectors.toarray()\n",
    "N_digrams = np.shape(stat_vectors)[1]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2340516a-4a70-45a0-81f3-ffe9c8d2fd5c",
   "metadata": {},
   "source": [
    "Now that we've vectorized the word data and the stats, we can combine them:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "6ec7b9f2-4c5d-4548-b533-34c5867f2b46",
   "metadata": {},
   "outputs": [],
   "source": [
    "vectors = np.concatenate((token_vectors, stat_vectors), axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "407ca5b6-aafa-48c8-a77f-1704f56ce103",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## Information Weighting\n",
    "Now a pokemon is a binary vector which encodes what moves, abilities, types and stats it has. However, not all moves and abilities are equally informative about what distinguishes pokemon. Some moves or abilities are very unique and define the pokemon that have them, such as Spore or Huge Power. Others are commonplace like Tackle. \n",
    "\n",
    "When measuring the distance between two pokemon we want to place a higher weight on these more distinctive moves. The Information Weight Transform takes a list of vectors and does exactly this; it weights each feature (move, ability, etc) by the amount of <i> information </i> that it tells us about the pokemon, relative to all other pokemon. The working principle is similar to TF-IDF, if you've heard of it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "05fc1ead-15af-44d4-b186-78b8e6246f38",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/kdruscit/.conda/envs/sklearn/lib/python3.12/site-packages/numba/np/ufunc/parallel.py:371: NumbaWarning: The TBB threading layer requires TBB version 2021 update 6 or later i.e., TBB_INTERFACE_VERSION >= 12060. Found TBB_INTERFACE_VERSION = 12050. The TBB threading layer is disabled.\n",
      "  warnings.warn(problem)\n"
     ]
    }
   ],
   "source": [
    "## Information Weighting:\n",
    "IWT = InformationWeightTransformer()\n",
    "vectors_IWT = IWT.fit_transform(vectors)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bc4ebb7a-9800-4c70-a8f0-4be4683e36ab",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Dimension Reduction: UMAP\n",
    "Now in order to categorize pokemon and plot them on a map, we need to do a dimension reduction. The vectors we've built have too many dimensions to visualize:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "ac0edc0e-43a7-4e08-b4dd-296c8fb73881",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The vectors have 1117 dimensions!\n"
     ]
    }
   ],
   "source": [
    "print(\"The vectors have\",np.shape(vectors)[1],\"dimensions!\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aebf1b93-5380-45e8-801e-237ccabe969a",
   "metadata": {},
   "source": [
    "Dimension reduction tools allow you to map a set of vectors in high dimensions to a lower dimensional space, while approximately preserving the distance between the vectors. Here again we use UMAP. UMAP has a few parameters we can select, see <a href=https://umap-learn.readthedocs.io/en/latest/parameters.html> the UMAP docs.</a> I encourage you to try re-running UMAP with different parameters to see how it affects the output.\n",
    "\n",
    "Inputting our information-weighted vectors, we will get out a pair of co-ordinates (x,y) for each pokemon, which we can put in a scatter plot."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "e9408ae3-562b-4880-98b3-76838b23b0a2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# This may take a few seconds.\n",
    "use_included_data = True\n",
    "\n",
    "if use_included_data:\n",
    "    ## This loads the data exactly as used on the website.\n",
    "    data = np.load(\"data/pokedex_umap.npy\")\n",
    "else:\n",
    "    ## This generates a new map, which will break the labels (see below).\n",
    "    UMAP2d = UMAP(n_components=2, metric='cosine')\n",
    "    data = UMAP2d.fit_transform(vectors_IWT)\n",
    "plt.scatter(\n",
    "    data[:,0],\n",
    "     data[:,1],\n",
    "     marker='.',\n",
    ") \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "954e66fe-70fa-4ff6-abd8-d822fdcdf4ec",
   "metadata": {},
   "source": [
    "There are a couple things to keep in mind: <br>\n",
    "First, the absolute x and y scales of the UMAP data do not have any meaning. Only the relative distance between points means anything. <br>\n",
    "Second, while UMAP preserves local distances fairly well, the long-range distances must be distorted. Compromises must be made to reduce 1000+ dimensions down to 2!"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9a3c05d2-07f3-4a5c-a2b9-c6c817167986",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Clustering\n",
    "Now that our Pokemon are represented as points in the plane, we can run a clustering algorithm to find dense groupings of pokemon. At the Tutte Institute we are partial to our inhouse algorithm, HDBSCAN.\n",
    "\n",
    "HDBSCAN will assign an integer cluster label to each point, indicating which cluster it belongs to. Points which get -1 are considered to be \"noise\", meaning they are outliers or they fit between multiple clusters. \n",
    "\n",
    "The critical parameter of HDBSCAN is the minimum cluster size, which you should set depending on how fine a resolution of topic you want to find. In my case, I want pretty fine clusters, but I don't want a single evolution family of Pokemon to be there own cluster (for example, Bulbasaur, Ivysaur, Venusaur). The largest Pokemon families have only 3 members, so I will set the minimum cluster size to 4."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "dc982bbf-b8cf-457c-a36f-e65e226fe76e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "146 noise points: 14.243902439024389 %\n"
     ]
    }
   ],
   "source": [
    "clusterer = HDBSCAN(min_cluster_size=4)\n",
    "clusterer.fit(data)\n",
    "cluster_labels = clusterer.labels_\n",
    "plt.scatter(\n",
    "    data[:,0],\n",
    "    data[:,1],\n",
    "    marker='.',\n",
    "    c=cluster_labels\n",
    ")\n",
    "plt.show()\n",
    "n_noise = np.size((cluster_labels==-1).nonzero())\n",
    "print(n_noise,\"noise points:\", (n_noise/1025)*100, \"%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b9574d7f-ec9b-4479-afbf-042902db80eb",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Labelling Clusters\n",
    "Unfortunately, HDBSCAN doesn't know anything about pokemon. Or about anything really. So it can only give the clusters a numeric label. To get more interesting labels, we need to take a look at the data points that lie within each cluster.\n",
    "\n",
    "To do this, we can look at the pokemon within each cluster, and we can also get some of the highest information-weighted keywords that are associated to each cluster."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "72094fc2-770a-4e60-b0ea-7baff43c66a7",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "## Get keywords & members of each cluster.\n",
    "N_clusters = np.size(np.unique(cluster_labels))\n",
    "N_features = np.shape(vectors)[1]\n",
    "cluster_vectors = np.zeros((N_clusters, N_features))\n",
    "keywords = np.zeros((1025,3), dtype=object)\n",
    "for i, c_name in enumerate(np.unique(cluster_labels)):\n",
    "    c_idx = (cluster_labels == c_name).nonzero()\n",
    "    mons_in_cluster = pokedex['name'].iloc[c_idx].values\n",
    "    cluster_sum_vector = np.sum(vectors_IWT[c_idx],axis=0)\n",
    "    cluster_vectors[i] = cluster_sum_vector\n",
    "    \n",
    "    \n",
    "weighted_cvecs = IWT.fit_transform(cluster_vectors)\n",
    "for i, c_name in enumerate(np.unique(cluster_labels)):\n",
    "    c_idx = (cluster_labels == c_name).nonzero()\n",
    "    srt = np.argsort(weighted_cvecs[i])\n",
    "    cl_keywords = np.array(['','',''],dtype=object)\n",
    "    for k in range(1,4):\n",
    "        if srt[-k] < N_features - N_digrams:\n",
    "            cl_keywords[k-1] = vectorizer.get_feature_names_out()[srt[-k]]\n",
    "        else:\n",
    "            cl_keywords[k-1] = stat_vectorizer.get_feature_names_out()[srt[-k]-(N_features - N_digrams)]\n",
    "    keywords[c_idx] = cl_keywords\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "0d28ae09-ae67-4eaa-bfd4-6e70b392cccd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CLUSTER: 33\n",
      "Keywords for this cluster are: ['synchronize' 'forewarn' 'mystical_power']\n",
      "Pokemon in this cluster are: ['abra' 'kadabra' 'alakazam' 'drowzee' 'hypno' 'starmie' 'mr-mime' 'jynx'\n",
      " 'mewtwo' 'natu' 'xatu' 'espeon' 'umbreon' 'unown' 'girafarig' 'stantler'\n",
      " 'smoochum' 'ralts' 'kirlia' 'gardevoir' 'lunatone' 'solrock' 'baltoy'\n",
      " 'claydol' 'chimecho' 'latias' 'latios' 'deoxys-normal' 'chingling'\n",
      " 'bronzor' 'bronzong' 'mime-jr' 'uxie' 'mesprit' 'azelf' 'cresselia'\n",
      " 'munna' 'musharna' 'solosis' 'duosion' 'reuniclus' 'elgyem' 'beheeyem'\n",
      " 'cosmoem' 'lunala' 'orbeetle' 'indeedee-male' 'eternatus' 'wyrdeer'\n",
      " 'rabsca' 'farigiraf']\n",
      "-------------------------\n"
     ]
    }
   ],
   "source": [
    "def clusterInfo(c_val):\n",
    "    cl_idx = (cluster_labels==c_val).nonzero()\n",
    "    print(\"CLUSTER:\",c_val)\n",
    "    print(\"Keywords for this cluster are:\",keywords[cl_idx[0][0]])\n",
    "    print(\"Pokemon in this cluster are:\", pokedex['name'].to_numpy()[cl_idx])\n",
    "    print(\"-------------------------\")\n",
    "    \n",
    "clusterInfo(33)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e0ccbccb-331c-413f-8cc8-110444272169",
   "metadata": {},
   "source": [
    "To turn this list of pokemon and keywords into a short name for the cluster, one could consider writing a prompt and asking an LLM to provide a distinctive name. This turns out to be a subtle problem; see https://github.com/tutteinstitute/topicnaming .\n",
    "Since we don't have that many clusters and I'm writing this tutorial with my dinky laptop over VIA Rail Wi-Fi, I will look at the keywords and pokemon and give them a name myself.\n",
    "\n",
    "If you have a better suggestion for any of these names, I'd love to hear it!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "74501439-b714-44d1-bada-a216bb55fde9",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "'''\n",
    "WARNING: If you re-ran the code then you will have different clusters (due to UMAP randomness) and the naming will not match. \n",
    "'''\n",
    "cluster_names = {\n",
    "    0:\"Future Paradox\",\n",
    "    1:\"Past Paradox\",\n",
    "    2:\"Ice\",\n",
    "    3:\"Fire\",\n",
    "    4:\"Fire Starters\",\n",
    "    5:\"Magic\",\n",
    "    6:\"DEF&HP\",\n",
    "    7:\"Applin\",\n",
    "    8:\"Birds\",\n",
    "    9:\"Mushrooms\",\n",
    "    10:\"Flower\",\n",
    "    11:\"Food Fairies\",\n",
    "    12:\"Phys. Grass Starters\",\n",
    "    13:\"Horn Leech\",\n",
    "    14:\"Fruit\",\n",
    "    15:\"Fangs\",\n",
    "    16:\"Rivalry\",\n",
    "    17:\"Spec. Grass Starters\",\n",
    "    18:\"Dragon\",\n",
    "    19:\"Utility Grass\",\n",
    "    20:\"Leaf/Vine\",\n",
    "    21:\"Jaws\",\n",
    "    22:\"Shed Skin\",\n",
    "    23:\"Weapons\",\n",
    "    24:\"Arachnids\",\n",
    "    25:\"Water Starters\",\n",
    "    26:\"Greninja\",\n",
    "    27:\"Lepidoptera 1\",\n",
    "    28:\"Lepidoptera 2\",\n",
    "    29:\"Fairy\",\n",
    "    30:\"???\",\n",
    "    31:\"Rodents\",\n",
    "    32:\"Shadow Tag\",\n",
    "    33:\"Psychic\",\n",
    "    34:\"Cephalopods\",\n",
    "    35:\"Hungry\",\n",
    "    36:\"Fighting\",\n",
    "    37:\"Rock\",\n",
    "    38:\"Ghost\",\n",
    "    39:\"Poison\",\n",
    "    40:\"Pinchers\",\n",
    "    41:\"Dogs\",\n",
    "    42:\"Mechanical\",\n",
    "    43:\"Electric\",\n",
    "    44:\"Sword\",\n",
    "    45:\"Fish\",\n",
    "    46:\"Shells\",\n",
    "    47:\"Poison Barbs\",\n",
    "    48:\"Ground/Rock\",\n",
    "    49:\"Bulky ?\",\n",
    "    50:\"Water\",\n",
    "    -1:\"Unclassified\"\n",
    "}\n",
    "\n",
    "try:\n",
    "    string_labels = [cluster_names[x] for x in cluster_labels]\n",
    "except(KeyError):\n",
    "    print(\"Clusters do not match labels!\")\n",
    "    string_labels = [''.join(x+\"\\n \") for x in keywords]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f7eb055f-5b73-4b58-889b-53cfc8143837",
   "metadata": {},
   "source": [
    "## Creating the Datamap\n",
    "Now all that is left to do is a make a nice looking plot! To do this we will use datamapplot. First we can generate some nice looking labels for each Pokemon when you hover over it's marker."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "90578b02-5c7a-4fc2-b418-cf4d6fcd1401",
   "metadata": {},
   "outputs": [],
   "source": [
    "hover_data = pd.DataFrame()\n",
    "names_nice = [x.capitalize() for x in pokedex['name'].to_list()]\n",
    "hover_data[\"Name\"]=names_nice\n",
    "type_nice = []\n",
    "for row in pokedex['type'].to_list():\n",
    "    tmp = ''\n",
    "    for word in row:\n",
    "        tmp += word.capitalize()+\" \"\n",
    "    type_nice.append(tmp)\n",
    "hover_data[\"Type\"]=type_nice\n",
    "hover_data[\"TopStats\"] = pokedex[\"top_stat\"]+\" \"+pokedex[\"second_stat\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "af3f31bc-81d6-4df3-8991-d4de57a97ca0",
   "metadata": {},
   "outputs": [],
   "source": [
    "distinctive_words = []\n",
    "for i, mon in enumerate(pokedex['name']):\n",
    "    vec = vectors_IWT[i]\n",
    "    srt = np.argsort(vec)\n",
    "    kw =  ''\n",
    "    for k in range(1,3):\n",
    "        if srt[-k] < N_features - N_digrams:\n",
    "            kw += vectorizer.get_feature_names_out()[srt[-k]].capitalize()\n",
    "            kw += \" \"\n",
    "        else:\n",
    "            kw += stat_vectorizer.get_feature_names_out()[srt[-k]-(N_features - N_digrams)].capitalize()\n",
    "            kw += \" \"\n",
    "    distinctive_words.append(kw)\n",
    "hover_data[\"KW\"] = distinctive_words"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "75cca9b1-d532-435c-a739-b10c3a08ac97",
   "metadata": {},
   "source": [
    "We can also resize each marker by the pokemon's BST."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "ef8fe902-7351-43a4-8de7-6a70e526ee5d",
   "metadata": {},
   "outputs": [],
   "source": [
    "size_arr = np.zeros(1025)\n",
    "for i, statline in enumerate(pokedex['stats']):\n",
    "    size_arr[i] = (np.sum(statline)/400)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "519210d6-b22a-4ac8-bb6b-3b3d6d9d275a",
   "metadata": {},
   "source": [
    " Lastly, we pass all the arguments to datamapplot!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "95606e4d-dc1b-4e56-bc49-09fba326ddf4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "            <iframe\n",
       "                width=100%\n",
       "                height=800\n",
       "                frameborder=\"0\"\n",
       "                srcdoc=\"\n",
       "&lt;!DOCTYPE html&gt;\n",
       "&lt;html&gt;\n",
       "  &lt;head&gt;\n",
       "    &lt;meta http-equiv=&quot;content-type&quot; content=&quot;text/html; charset=UTF-8&quot; /&gt;\n",
       "    &lt;title&gt;Pok&amp;eacute;Map&lt;/title&gt;\n",
       "    \n",
       "    &lt;link rel=&quot;preconnect&quot; href=&quot;https://fonts.googleapis.com&quot;&gt;\n",
       "    &lt;link rel=&quot;preconnect&quot; href=&quot;https://fonts.gstatic.com&quot; crossorigin&gt;\n",
       "    &lt;link href=&quot;https://fonts.googleapis.com/css2?family=Roboto&amp;display=swap&quot; rel=&quot;stylesheet&quot;&gt;\n",
       "    \n",
       "    \n",
       "       \n",
       "    &lt;link rel=&quot;stylesheet&quot; href=&quot;https://maxcdn.bootstrapcdn.com/bootstrap/3.2.0/css/bootstrap-theme.min.css&quot; /&gt;\n",
       "    &lt;link rel=&quot;stylesheet&quot; href=&quot;https://maxcdn.bootstrapcdn.com/font-awesome/4.6.3/css/font-awesome.min.css&quot; /&gt;\n",
       "    &lt;script src=&quot;https://unpkg.com/deck.gl@latest/dist.min.js&quot;&gt;&lt;/script&gt;\n",
       "    \n",
       "    &lt;script src=&quot;https://unpkg.com/fflate@0.8.0&quot;&gt;&lt;/script&gt;\n",
       "    \n",
       "    &lt;style&gt;\n",
       "        body {\n",
       "          margin: 0;\n",
       "          padding: 0;\n",
       "          overflow: hidden;\n",
       "          background: #ffffff;\n",
       "        }\n",
       "\n",
       "        #deck-container {\n",
       "          width: 100vw;\n",
       "          height: 100vh;\n",
       "        }\n",
       "\n",
       "        #deck-container canvas {\n",
       "          z-index: 1;\n",
       "          background: #ffffff;\n",
       "        }\n",
       "\n",
       "        .deck-tooltip {\n",
       "            \n",
       "            font-size: 0.8em;\n",
       "            font-family: Roboto;\n",
       "            font-weight: 300;\n",
       "            color: #000000 !important;\n",
       "            background-color: #ffffffaa !important;\n",
       "            border-radius: 12px;\n",
       "            box-shadow: 2px 3px 10px #aaaaaa44;\n",
       "            max-width: 25%;\n",
       "        }\n",
       "        \n",
       "        #loading {\n",
       "            width: 100%;\n",
       "            height: 100%;\n",
       "            top: 0px;\n",
       "            left: 0px;\n",
       "            position: absolute;\n",
       "            display: block; \n",
       "            z-index: 99\n",
       "        }\n",
       "\n",
       "        #loading-image {\n",
       "            position: absolute;\n",
       "            top: 45%;\n",
       "            left: 47.5%;\n",
       "            z-index: 100\n",
       "        }\n",
       "        \n",
       "        #title-container {\n",
       "            position: absolute;\n",
       "            top: 0;\n",
       "            left: 0;\n",
       "            margin: 16px;\n",
       "            padding: 12px;\n",
       "            border-radius: 16px;\n",
       "            line-height: 0.95;\n",
       "            z-index: 2;\n",
       "            font-family: Roboto;\n",
       "            color: #000000;\n",
       "            background: #ffffffaa;\n",
       "            box-shadow: 2px 3px 10px #aaaaaa44;\n",
       "        }\n",
       "        \n",
       "        \n",
       "        #search-container{\n",
       "            position: absolute;\n",
       "            left: -16px;\n",
       "            margin: 16px;\n",
       "            padding: 12px;\n",
       "            border-radius: 16px;\n",
       "            z-index: 2;\n",
       "            font-family: Roboto;\n",
       "            color: #000000;\n",
       "            background: #ffffffaa;\n",
       "            width: fit-content;\n",
       "            box-shadow: 2px 3px 10px #aaaaaa44;\n",
       "        }\n",
       "        input {\n",
       "            margin: 2px;\n",
       "            padding: 4px;\n",
       "            border-radius: 8px;\n",
       "            color: #000000;\n",
       "            background: #ffffffdd;\n",
       "            border: 1px solid #ddddddff;\n",
       "            transition: 0.5s;\n",
       "            outline: none;\n",
       "        }\n",
       "        input:focus {\n",
       "          border: 2px solid #555;\n",
       "        }\n",
       "        \n",
       "        \n",
       "    &lt;/style&gt;\n",
       "  &lt;/head&gt;\n",
       "  &lt;body&gt;\n",
       "    &lt;div id=&quot;loading&quot;&gt;\n",
       "        &lt;img id=&quot;loading-image&quot; src=&quot;https://i.gifer.com/ZKZg.gif&quot; alt=&quot;Loading...&quot; width=&quot;5%&quot;/&gt;\n",
       "    &lt;/div&gt;\n",
       "    \n",
       "    &lt;div id=&quot;title-container&quot;&gt;\n",
       "        &lt;span style=&quot;font-family:Roboto;font-size:36pt;color:#000000&quot;&gt;\n",
       "            Pok&amp;eacute;Map\n",
       "        &lt;/span&gt;&lt;br/&gt;\n",
       "        &lt;span style=&quot;font-family:Roboto;font-size:18pt;color:#777777&quot;&gt;\n",
       "            \n",
       "        &lt;/span&gt;\n",
       "        \n",
       "        &lt;div id=&quot;search-container&quot;&gt;\n",
       "            &lt;input autocomplete=&quot;off&quot; type=&quot;search&quot; id=&quot;search&quot; placeholder=&quot;🔍&quot;&gt;\n",
       "        &lt;/div&gt;\n",
       "        \n",
       "    &lt;/div&gt;\n",
       "    \n",
       "    \n",
       "    &lt;div id=&quot;deck-container&quot;&gt;\n",
       "    &lt;/div&gt;\n",
       "    \n",
       "  &lt;/body&gt;\n",
       "  &lt;script type=&quot;module&quot;&gt;\n",
       "    import { ArrowLoader } from &#x27;https://cdn.jsdelivr.net/npm/@loaders.gl/arrow@4.1.0-alpha.10/+esm&#x27;\n",
       "    import { JSONLoader } from &#x27;https://cdn.jsdelivr.net/npm/@loaders.gl/json@4.0.5/+esm&#x27;\n",
       "    \n",
       "\n",
       "    \n",
       "    const pointDataBase64 = &quot;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&quot;;\n",
       "    const pointDataBuffer = fflate.strToU8(atob(pointDataBase64), true);\n",
       "    const pointData = await loaders.parse(pointDataBuffer, ArrowLoader);\n",
       "    const hoverDataBase64 = &quot;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&quot;;\n",
       "    const hoverDataBuffer = fflate.strToU8(atob(hoverDataBase64), true);\n",
       "    const unzippedHoverData = fflate.gunzipSync(hoverDataBuffer);\n",
       "    const hoverData = await loaders.parse(unzippedHoverData, ArrowLoader);\n",
       "    const labelDataBase64 = &quot;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&quot;;\n",
       "    const labelDataBuffer = fflate.strToU8(atob(labelDataBase64), true);\n",
       "    const unzippedLabelData = fflate.gunzipSync(labelDataBuffer);    \n",
       "    const labelData = await loaders.parse(unzippedLabelData, JSONLoader);\n",
       "    \n",
       "    \n",
       "    const DATA = {src: pointData.data, length: pointData.data.x.length}\n",
       "\n",
       "    const container = document.getElementById(&#x27;deck-container&#x27;);\n",
       "    const pointLayer = new deck.ScatterplotLayer({\n",
       "        id: &#x27;dataPointLayer&#x27;,\n",
       "        data: DATA,\n",
       "        getPosition: (object, {index, data}) =&gt; {\n",
       "            return [data.src.x[index], data.src.y[index]];\n",
       "        },\n",
       "        \n",
       "        getRadius: (object, {index, data}) =&gt; {\n",
       "            return data.src.size[index];\n",
       "        },\n",
       "        \n",
       "        getFillColor: (object, {index, data}) =&gt; {\n",
       "            return [\n",
       "                data.src.r[index], \n",
       "                data.src.g[index], \n",
       "                data.src.b[index],\n",
       "                180\n",
       "            ]\n",
       "        },\n",
       "        getLineColor: (object, {index, data}) =&gt; {\n",
       "            return [\n",
       "                data.src.r[index], \n",
       "                data.src.g[index], \n",
       "                data.src.b[index],\n",
       "                32\n",
       "            ]\n",
       "        },       \n",
       "        getLineColor: [250, 250, 250, 128],\n",
       "        getLineWidth: 0.001,\n",
       "        highlightColor: [170, 0, 0, 187], \n",
       "        lineWidthMaxPixels: 8,\n",
       "        lineWidthMinPixels: 0.1,\n",
       "        radiusMaxPixels: 24, \n",
       "        radiusMinPixels: 0.01,\n",
       "        radiusUnits: &quot;common&quot;, \n",
       "        lineWidthUnits: &quot;common&quot;, \n",
       "        autoHighlight: true,\n",
       "        pickable: true, \n",
       "        stroked: true\n",
       "    });\n",
       "    const labelLayer = new deck.TextLayer({\n",
       "        id: &quot;textLabelLayer&quot;,\n",
       "        data: labelData,\n",
       "        pickable: false,\n",
       "        getPosition: d =&gt; [d.x, d.y],\n",
       "        getText: d =&gt; d.label,\n",
       "        getColor: d =&gt; [d.r, d.g, d.b],\n",
       "        getSize: d =&gt; d.size,\n",
       "        sizeScale: 1,\n",
       "        sizeMinPixels: 18,\n",
       "        sizeMaxPixels: 36,\n",
       "        outlineWidth: 8,\n",
       "        outlineColor: [238, 238, 238, 221],\n",
       "        getBackgroundColor: [255, 255, 255, 64],\n",
       "        getBackgroundPadding: [15, 15, 15, 15],\n",
       "        background: true,\n",
       "        characterSet: &quot;auto&quot;,\n",
       "        fontFamily: &quot;Roboto&quot;,\n",
       "        fontWeight: 900,\n",
       "        lineHeight: 0.95,\n",
       "        fontSettings: {&quot;sdf&quot;: true},\n",
       "        getTextAnchor: &quot;middle&quot;,\n",
       "        getAlignmentBaseline: &quot;center&quot;,\n",
       "        lineHeight: 0.95,\n",
       "        elevation: 100,\n",
       "        // CollideExtension options\n",
       "        collisionEnabled: true,\n",
       "        getCollisionPriority: d =&gt; d.size,\n",
       "        collisionTestProps: {\n",
       "          sizeScale: 3,\n",
       "          sizeMaxPixels: 36 * 2,\n",
       "          sizeMinPixels: 18 * 2\n",
       "        },\n",
       "        extensions: [new deck.CollisionFilterExtension()],\n",
       "    });\n",
       "    \n",
       "\n",
       "    const deckgl = new deck.DeckGL({\n",
       "      container: container,\n",
       "      initialViewState: {\n",
       "        latitude: 8.601964,\n",
       "        longitude: 10.077544,\n",
       "        zoom: 5.085960600376048\n",
       "      },\n",
       "      controller: true,\n",
       "      \n",
       "      layers: [pointLayer, labelLayer],\n",
       "      \n",
       "      \n",
       "      getTooltip: ({index, picked}) =&gt; picked ? {&quot;html&quot;: `\n",
       "        &lt;div&gt;\n",
       "            &lt;h3&gt;${hoverData.data.hover_text[index]}&lt;/h3&gt;\n",
       "            &lt;p&gt;Type: ${hoverData.data.Type[index]}&lt;/p&gt;\n",
       "            &lt;p&gt;Top Stats: ${hoverData.data.TopStats[index]}&lt;/p&gt;\n",
       "            &lt;p&gt;Notable: ${hoverData.data.KW[index]}&lt;/p&gt;\n",
       "        &lt;/div&gt;\n",
       "        `} : null\n",
       "    });\n",
       "    \n",
       "    document.getElementById(&quot;loading&quot;).style.display = &quot;none&quot;;\n",
       "        \n",
       "    \n",
       "        function selectPoints(item, conditional) {\n",
       "        var layerId;\n",
       "        if (item) {\n",
       "            for (var i = 0; i &lt; DATA.length; i++) {\n",
       "                if (conditional(i)) {\n",
       "                    DATA.src.selected[i] = 1;\n",
       "                } else {\n",
       "                    DATA.src.selected[i] = 0;\n",
       "                }\n",
       "            }\n",
       "            layerId = &#x27;selectedPointLayer&#x27; + item;\n",
       "        } else {\n",
       "            for (var i = 0; i &lt; DATA.length; i++) {\n",
       "                DATA.src.selected[i] = 1;\n",
       "            }\n",
       "            layerId = &#x27;dataPointLayer&#x27;;\n",
       "        }\n",
       "        const selectedPointLayer = pointLayer.clone(\n",
       "            {\n",
       "                id: layerId,\n",
       "                data: DATA,\n",
       "                getFilterValue: (object, {index, data}) =&gt; data.src.selected[index],\n",
       "                filterRange: [1, 2],\n",
       "                extensions: [new deck.DataFilterExtension({filterSize: 1})]\n",
       "            }\n",
       "        );\n",
       "        deckgl.setProps(\n",
       "            {layers: \n",
       "                [selectedPointLayer].concat(deckgl.props.layers.slice(1,))\n",
       "            }\n",
       "        );\n",
       "    }\n",
       "    \n",
       "    const search = document.getElementById(&quot;search&quot;);\n",
       "    search.addEventListener(&quot;input&quot;, (event) =&gt; {\n",
       "            const search_term = event.target.value.toLowerCase();\n",
       "            selectPoints(search_term, (i) =&gt; hoverData.data.hover_text[i].toLowerCase().includes(search_term));\n",
       "        }\n",
       "    );\n",
       "    \n",
       "    \n",
       "    &lt;/script&gt;\n",
       "&lt;/html&gt;\"\n",
       "            ></iframe>\n",
       "        "
      ],
      "text/plain": [
       "<InteractiveFigure width=100% height=800>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fig = datamapplot.create_interactive_plot(\n",
    "    data,\n",
    "    string_labels,\n",
    "    title=\"Pok&eacute;Map\",\n",
    "    hover_text=hover_data['Name'].to_list(),\n",
    "    extra_point_data = hover_data,\n",
    "    hover_text_html_template = \"\"\"\n",
    "        <div>\n",
    "            <h3>{hover_text}</h3>\n",
    "            <p>Type: {Type}</p>\n",
    "            <p>Top Stats: {TopStats}</p>\n",
    "            <p>Notable: {KW}</p>\n",
    "        </div>\n",
    "        \"\"\",\n",
    "    enable_search=True,\n",
    "    marker_size_array=size_arr,\n",
    "    #font_family=\"Arial\",\n",
    ")\n",
    "fig"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "43c69270-4175-49ad-aba3-f518b9dfdef7",
   "metadata": {},
   "source": [
    "## Conclusion\n",
    "In this tutorial, we introduced topic modelling. In particular I've tried to convey the type of considerations that go into into making sensible models of your data. Try running the pipeline again with different choices - you'll see that some groups of Pokemon almost always appear, such as the Paradox pokemon, but other groups are much more ephermeral. \n",
    "\n",
    "My interpretation is that any conclusion one obtains with a set of choices is a piece of evidence for the truth of the underlying statement. If many different sets of choices lead to the same conclusion, that means there is a lot of evidence that it is true! If a conclusion only occurs under a very specific set of choices, then those choices much be very strongly justified, otherwise the conclusion is probably spurrious."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1cf6a9bf-4a2f-420e-963c-54db1f5e9809",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python (sklearn)",
   "language": "python",
   "name": "myenv"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}