this_algorithm/wordlist/wordlist-new2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "991a711f-be98-4aae-a657-84b065449916",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "try:\n",
    "    _initialized\n",
    "except:\n",
    "    # !pip install spacy\n",
    "    # !python -m spacy download en_core_web_trf\n",
    "    import spacy\n",
    "    \n",
    "    nlp = spacy.load('en_core_web_trf', disable=['parser', 'ner'])\n",
    "    \n",
    "    _initialized=True\n",
    "    \n",
    "import pandas as pd\n",
    "import gzip\n",
    "import re"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "d130bb84",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_lines(filename):\n",
    "    with gzip.open(filename, 'r') as f:\n",
    "        ret = []\n",
    "        for l in f:\n",
    "            if len(ret) > 30_000:\n",
    "                return ret\n",
    "            ret.append(str(l).lower())\n",
    "        return ret\n",
    "\n",
    "\n",
    "    \n",
    "WORDLIST_SIZE = 8192 + 3\n",
    "word_re = re.compile(r\"^[A-Za-z]+$\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "de2d1731",
   "metadata": {},
   "outputs": [],
   "source": [
    "!pwd\n",
    "!ls"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "90665714",
   "metadata": {},
   "outputs": [],
   "source": [
    "annotated_words=pd.read_excel(\"annotated_words.ods\")\n",
    "\n",
    "excluded_words = list(annotated_words[annotated_words[\"keep\"] != \"Yes\"][\"word\"].str.lower())\n",
    "excluded_words[0:10]\n",
    "\n",
    "custom_maps = annotated_words[annotated_words[\"maps_to\"].notna()][[\"word\",\"maps_to\"]].assign(maps_to=lambda x: x[\"maps_to\"].map(lambda y: y.split(\",\")))\n",
    "\n",
    "custom_maps = [\n",
    "    (m[1][\"word\"].lower(), mapping.lower())\n",
    "    for m in custom_maps.iterrows()\n",
    "    for mapping in m[1][\"maps_to\"]\n",
    "]\n",
    "custom_maps"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "fb50c69e",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Start parsing the wordlist\n",
    "all_words = get_lines(\"00-frequency-all.txt.gz\")\n",
    "\n",
    "# Delete header line\n",
    "all_words = all_words[1:]\n",
    "\n",
    "# Get only the word (fixed width)\n",
    "all_words = [w[13:36].strip() for w in all_words]\n",
    "\n",
    "# Remove special characters\n",
    "all_words = [w for w in all_words if word_re.search(w)]\n",
    "\n",
    "# Remove all removed words\n",
    "all_words = [w for w in all_words if w not in excluded_words]\n",
    "\n",
    "# Add all custom mappings\n",
    "for m in list(sum(custom_maps, ())):\n",
    "    if m[0] not in all_words:\n",
    "        all_words.append(m[0])\n",
    "    if m[1] not in all_words:\n",
    "        all_words.append(m[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "cd21bff5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Lemmatize all words (plural -> singular)\n",
    "lemmatize_mappings = [(w, nlp(w)[0].lemma_) for w in all_words[:100]]\n",
    "print(lemmatize_mappings[:100])\n",
    "\n",
    "# Add custom lemmatizations\n",
    "for l in custom_maps:\n",
    "    if l in lemmatize_mappings:\n",
    "        print(f\"Warning: {l} is already lemmatized\")\n",
    "    else:\n",
    "        lemmatize_mappings.append(l)\n",
    "        \n",
    "print(lemmatize_mappings[:100])\n",
    "\n",
    "lemmatize_mappings = [w for w in lemmatize_mappings if w[1] not in excluded_words]\n",
    "print(lemmatize_mappings[:100])\n",
    "\n",
    "# Now, re-add all lematized words to the list of every word\n",
    "for w in sum(lemmatize_mappings, ()):\n",
    "    if w not in all_words:\n",
    "        print(w)\n",
    "        all_words.append(w)\n",
    "        \n",
    "lemmatize_mappings = {k: v for k, v in lemmatize_mappings}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "0ee9af7d",
   "metadata": {},
   "outputs": [],
   "source": [
    "final_wordlist = []\n",
    "seen_lemmatizations = set()\n",
    "for w in all_words:\n",
    "    lemmatized = lemmatize_mappings.get(w) or w\n",
    "    if lemmatized in seen_lemmatizations:\n",
    "        # The lemmatized version of this word was already seen\n",
    "        continue\n",
    "    else:\n",
    "        # The lemmatized version hasn't been seen. We're good to add it\n",
    "        final_wordlist.append([\n",
    "            k\n",
    "            for k\n",
    "            in lemmatize_mappings.keys()\n",
    "            if lemmatize_mappings[k] == lemmatized\n",
    "        ])\n",
    "        seen_lemmatizations.add(lemmatized)\n",
    "\n",
    "    if len(final_wordlist) >= WORDLIST_SIZE:\n",
    "        break\n",
    "\n",
    "# Now, convert it to the format (number, word)\n",
    "final_wordlist = [\n",
    "    (idx, w)\n",
    "    for idx, words in enumerate(final_wordlist)\n",
    "    for w in words\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "07c1293c",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(len(lemmatize_mappings))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "19c255d0",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}