distance.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Processing social distancing tweets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "import datetime\n",
    "import os\n",
    "import re\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "DATADIR = \"/home/erikt/projects/puregome/data/distance/\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Frequencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "FILEPATTERN = \"20200[2-6]\"\n",
    "DATEPATTERN = \"%Y%m%d\"\n",
    "NBROFCHARSINDATE = 8"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "inFileNames = sorted(os.listdir(DATADIR))\n",
    "frequencies = {}\n",
    "for inFileName in inFileNames:\n",
    "    if re.search(FILEPATTERN,inFileName):\n",
    "        df = pd.read_csv(DATADIR+inFileName,compression=\"gzip\")\n",
    "        date = inFileName[:NBROFCHARSINDATE]\n",
    "        if not date in frequencies: frequencies[date] = 0\n",
    "        frequencies[date] += len(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "PEAK1 = \"20200322\"\n",
    "PEAK1TEXT = \" reactions on announcement\"\n",
    "PEAK2 = \"20200601\"\n",
    "PEAK2TEXT = \" demonstration at the Dam\"\n",
    "PEAK3 = \"20200315\"\n",
    "PEAK3OFFSET = \"20200206\"\n",
    "PEAK3TEXT = \" announcement of distancing\"\n",
    "\n",
    "def string2date(dateString,datePattern=DATEPATTERN):\n",
    "    return(datetime.datetime.strptime(dateString,datePattern))\n",
    "\n",
    "plt.figure(figsize=(12,6))\n",
    "ax1 = plt.subplot(111)\n",
    "ax1.plot_date([string2date(d) for d in frequencies],\\\n",
    "              [frequencies[d] for d in frequencies],xdate=True,fmt=\"-\",label=\"label\")\n",
    "plt.annotate(PEAK1TEXT,(string2date(PEAK1),frequencies[PEAK1]))\n",
    "plt.annotate(PEAK2TEXT,(string2date(PEAK2),frequencies[PEAK2]))\n",
    "plt.annotate(PEAK3TEXT,(string2date(PEAK3OFFSET),frequencies[PEAK3]))\n",
    "plt.title(\"Number of Dutch social distance tweets per day\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('20200322', 36538),\n",
       " ('20200601', 29775),\n",
       " ('20200321', 21488),\n",
       " ('20200602', 17530),\n",
       " ('20200323', 16900)]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[(d,frequencies[d]) for d in sorted(frequencies.keys(),key=lambda d:frequencies[d],reverse=True)][0:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Vocabulary comparison"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "import math\n",
    "import sys\n",
    "from IPython.display import clear_output\n",
    "from nltk.tokenize import TweetTokenizer\n",
    "sys.path.append(\"/home/erikt/projects/newsgac/fasttext-runs\")\n",
    "import tscore"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "ID = \"id_str\"\n",
    "TEXT = \"text\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "def squeal(text=None):\n",
    "    clear_output(wait=True)\n",
    "    if not text is None: print(text)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def getTokens(filePattern,dataDir=DATADIR):\n",
    "    fileList = sorted(os.listdir(dataDir))\n",
    "    tokens = {}\n",
    "    for inFileName in fileList:\n",
    "        if re.search(filePattern,inFileName):\n",
    "            squeal(inFileName)\n",
    "            df = pd.read_csv(DATADIR+inFileName,compression=\"gzip\",index_col=ID)\n",
    "            for i in range(0,len(df)):\n",
    "                text = df.iloc[i][TEXT]\n",
    "                for token in TweetTokenizer().tokenize(text.lower()): \n",
    "                    if not token in tokens: tokens[token] = 0\n",
    "                    tokens[token] += 1\n",
    "    return(tokens)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "NBROFTOKENS = \"totalFreq\"\n",
    "NBROFTYPES = \"nbrOfWords\"\n",
    "WORDFREQS = \"wordFreqs\"\n",
    "\n",
    "def makeTscoreData(tokenList):\n",
    "    data = { NBROFTOKENS:0, NBROFTYPES:0, WORDFREQS:{} }\n",
    "    for token in tokenList:\n",
    "        if not math.isnan(tokenList[token]):\n",
    "            data[WORDFREQS][token] = tokenList[token]\n",
    "            data[NBROFTYPES] += 1\n",
    "            data[NBROFTOKENS] += tokenList[token]\n",
    "    return(data)\n",
    "\n",
    "def sortTscores(tscores):\n",
    "    return({token:tscores[token] for token in sorted(tscores.keys(),key=lambda t:tscores[t],reverse=True)})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "NBROFEXAMPLES = 20\n",
    "\n",
    "def dictTopN(dictionary,n=NBROFEXAMPLES):\n",
    "    return([(x[1],x[0]) for x in dictionary.items()][0:n])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "20200601-23.out.gz\n"
     ]
    }
   ],
   "source": [
    "tokens1 = getTokens(PEAK1)\n",
    "tokens2 = getTokens(PEAK2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "93.9 thuis\n",
      "83.8 blijf\n",
      "73.7 houd\n",
      "72.9 afstand\n",
      "72.2 :\n",
      "60.3 het\n",
      "58.5 #coronavirus\n",
      "57.4 #houdafstand\n",
      "54.0 instructies\n",
      "52.9 of\n"
     ]
    }
   ],
   "source": [
    "tscoresData1 = makeTscoreData(tokens1)\n",
    "tscoresData2 = makeTscoreData(tokens2)\n",
    "for x in dictTopN(sortTscores(tscore.computeTscore(tscoresData1,tscoresData2)),n=10): \n",
    "    print(round(x[0],1),x[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "75.2 #halsema\n",
      "71.2 de\n",
      "70.9 dam\n",
      "68.0 1.5\n",
      "67.6 ben\n",
      "67.2 demonstratie\n",
      "58.3 halsema\n",
      "58.2 boete\n",
      "55.7 #amsterdam\n",
      "53.6 boven\n"
     ]
    }
   ],
   "source": [
    "for x in dictTopN(sortTscores(tscore.computeTscore(tscoresData2,tscoresData1)),n=10): \n",
    "    print(round(x[0],1),x[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}