annotated-graphs.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Annotated Graphs\n",
    "\n",
    "Make a graph for annotated data. The annotated data sets do not include time meta data so these need to be extracted of the json files.\n",
    "\n",
    "See also the notebook annotation-graphs, which provides a more efficient implementation for hour-level time retrieval."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Read annotated data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "import pandas as pd\n",
    "import re"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "TOPIC = \"blij\"\n",
    "TOPIC_EN = {\"blij\": \"happiness\", \"eenzaam\": \"loneliness\"}[TOPIC]\n",
    "\n",
    "if TOPIC == \"distance\":\n",
    "    DATA_FILE = f\"../{TOPIC}-tweets.csv\"\n",
    "elif TOPIC == \"mondkapje\":\n",
    "    DATA_FILE = f\"{TOPIC}-202003-202103-1000.csv\"\n",
    "elif TOPIC == \"testing\":\n",
    "    DATA_FILE = f\"{TOPIC}-202003-202012.csv\"\n",
    "elif TOPIC == \"vaccin\":\n",
    "    DATA_FILE = f\"{TOPIC}-202001-202101-1000.csv\"\n",
    "elif TOPIC == \"eenzaam\" or TOPIC == \"blij\":\n",
    "    DATA_FILE = f\"{TOPIC}_2020.csv\"\n",
    "else:\n",
    "    assert False, f\"unknown value for TOPIC: {TOPIC}\"\n",
    "DATA_DIR = \"../data/annotation/\"\n",
    "LABEL_FILE = DATA_FILE + \".human-labels.txt\"\n",
    "data_file_no_directory = DATA_FILE.split(\"/\")[-1]\n",
    "TIME_FILE = \"csv/time_data.\" + data_file_no_directory"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_csv(DATA_DIR + DATA_FILE)\n",
    "data.sort_values(by=['id_str'])\n",
    "label_data_raw = pd.read_csv(DATA_DIR + LABEL_FILE, header=None, sep=\" \")\n",
    "label_data_raw.sort_values(by=[2])\n",
    "\n",
    "label_data = {}\n",
    "for row in label_data_raw.iterrows():\n",
    "    annotator = row[1][0]\n",
    "    id_str = row[1][2]\n",
    "    label = row[1][4]\n",
    "    if annotator == \"erikt@xs4all.nl\":\n",
    "        label_data[id_str] = label"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Checks and data write (skip)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(620, 600)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(data), len(label_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.Series(label_data).value_counts()/len(label_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "data.iloc[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_doccano_data(data, label_data):\n",
    "    doccano_data = []\n",
    "    for i, row in data.iterrows():\n",
    "        if row[\"id_str\"] in label_data:\n",
    "            doccano_data.append({ \"id\": row[\"id_str\"], \"label\": label_data[row[\"id_str\"]], \"text\": re.sub(r\"\\\\n\", \" \", row[\"text\"]) })\n",
    "    return doccano_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def write_doccano_data(doccano_data, file_name):\n",
    "    out_file = open(file_name, \"w\")\n",
    "    for data in doccano_data:\n",
    "        print(json.dumps(data), file=out_file)\n",
    "    out_file.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "write_doccano_data(make_doccano_data(data, label_data), f\"doccano_{TOPIC}.json\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.DataFrame(make_doccano_data(data, label_data)).to_csv(f\"{TOPIC}.csv\", index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Get time meta data (skip)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import gzip\n",
    "import json\n",
    "import numpy as np\n",
    "import os\n",
    "import re\n",
    "from IPython.display import clear_output"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "JSON_DIR = \"/home/erikt/media/20190525/files/cloud/twitter/2020\"\n",
    "FIRST_TWEET_IDS_FILE = \"csv/first_tweet_ids.csv\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "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": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "if False:\n",
    "    first_tweet_ids = {}\n",
    "    for month in range(1, 3+1):\n",
    "        squeal(month)\n",
    "        month_dir = JSON_DIR + str(month).zfill(2) + \"/\"\n",
    "        file_names = sorted(os.listdir(month_dir))\n",
    "        for file_name in file_names:\n",
    "            infile = gzip.open(month_dir + file_name)\n",
    "            line = infile.readline()\n",
    "            infile.close()\n",
    "            json_data = json.loads(line)\n",
    "            first_tweet_ids[month_dir + file_name] = json_data['id_str']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "if False:\n",
    "    pd.DataFrame([{\"file_name\": file_name, \"id_str\": first_tweet_ids[file_name]} for file_name in first_tweet_ids]).to_csv(FIRST_TWEET_IDS_FILE, index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "first_tweet_ids = pd.read_csv(FIRST_TWEET_IDS_FILE).to_dict('records')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "first_tweet_ids_pointer = 0\n",
    "try:\n",
    "    time_data = pd.read_csv(TIME_FILE, index_col=\"id_str\").to_dict(orient=\"series\")['created_at']\n",
    "except:\n",
    "    time_data = {}\n",
    "for row in data.iterrows():\n",
    "    id_str = row[1]['id_str']\n",
    "    if id_str in label_data.keys() and id_str not in time_data:\n",
    "        id_str_found = False\n",
    "        while first_tweet_ids[first_tweet_ids_pointer+1]['id_str'] < id_str and \\\n",
    "              first_tweet_ids_pointer < len(first_tweet_ids)-1:\n",
    "            first_tweet_ids_pointer += 1\n",
    "        while first_tweet_ids[first_tweet_ids_pointer]['id_str'] > id_str and first_tweet_ids_pointer > 0:\n",
    "            first_tweet_ids_pointer -= 1\n",
    "        #print(first_tweet_ids[first_tweet_ids_pointer]['file_name'])\n",
    "        #print(id_str, first_tweet_ids[first_tweet_ids_pointer]['file_name'].split(\"/\")[-1].split(\".\")[0])\n",
    "        time_data[id_str] = first_tweet_ids[first_tweet_ids_pointer]['file_name'].split(\"/\")[-1].split(\".\")[0]\n",
    "        #infile = gzip.open(first_tweet_ids[first_tweet_ids_pointer]['file_name'])\n",
    "        #for line in infile:\n",
    "        #    json_data = json.loads(line)\n",
    "        #    if json_data['id_str'] == str(id_str):\n",
    "        #        time_data[json_data['id_str']] = json_data['created_at']\n",
    "        #        squeal(json_data['created_at'])\n",
    "        #        id_str_found = True\n",
    "        #        break\n",
    "        #infile.close()\n",
    "        #if not id_str_found and first_tweet_ids_pointer > 0:\n",
    "        #    first_tweet_ids_pointer -= 1\n",
    "        #    infile = gzip.open(first_tweet_ids[first_tweet_ids_pointer]['file_name'])\n",
    "        #    for line in infile:\n",
    "        #        json_data = json.loads(line)\n",
    "        #        if json_data['id_str'] == str(id_str):\n",
    "        #            time_data[json_data['id_str']] = json_data['created_at']\n",
    "        #            squeal(json_data['created_at'])\n",
    "        #            id_str_found = True\n",
    "        #            break\n",
    "        #    infile.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def file_date_to_json_date(file_date):\n",
    "    return datetime.datetime.strftime(datetime.datetime.strptime(file_date, \"%Y%m%d-%H\"), \"%a %b %d %H:%M:%S +0000 %Y\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "len(time_data.keys())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.DataFrame([{'id_str': id_str, 'created_at': file_date_to_json_date(time_data[id_str])} \n",
    "              for id_str in time_data.keys()]).to_csv(\"csv/time_data.\" + data_file_no_directory, index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Make graph"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import datetime\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.dates as mdates\n",
    "import matplotlib"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_file_no_directory = DATA_FILE.split(\"/\")[-1]\n",
    "TIME_FILE = \"csv/time_data.\" + data_file_no_directory"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "time_data = pd.read_csv(TIME_FILE, index_col=\"id_str\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_graph_data_list(label_data, time_data):\n",
    "    graph_data_dict = {}\n",
    "    for id_str in label_data:\n",
    "        label = label_data[id_str]\n",
    "        if label == \"EENS\" or label == \"ONEENS\" or label == \"RELEVANT\":\n",
    "            if id_str in time_data.index:\n",
    "                graph_data_dict[id_str] = { \"label\": label, \"created_at\": time_data.loc[id_str]['created_at'] }\n",
    "        elif label == \"ANDERS\" or label == \"NEUTRAL\" or label == \"IRRELEVANT\":\n",
    "            if id_str in graph_data_dict:\n",
    "                del(graph_data_dict[id_str])\n",
    "        else:\n",
    "            print(f\"make_graph_data: cannot happen: {label}\")\n",
    "    return([{'label': graph_data_dict[id_str]['label'], 'created_at': graph_data_dict[id_str]['created_at']} for id_str in sorted(graph_data_dict.keys())])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "N = 100\n",
    "\n",
    "\n",
    "def percentage(window):\n",
    "    total_count = 0\n",
    "    support_count = 0\n",
    "    for row in window:\n",
    "        total_count += 1\n",
    "        if row['label'] == \"EENS\":\n",
    "            support_count += 1\n",
    "    return(100 * support_count / total_count)\n",
    "\n",
    "\n",
    "def select_graph_data(graph_data_list):\n",
    "    window = []\n",
    "    x = []\n",
    "    y = []\n",
    "    for i in range(0, len(graph_data_list)):\n",
    "        window.append(graph_data_list[i])\n",
    "        if len(window) >= N:\n",
    "            selected_time_index = int((N-1)*(i-N+1)/(len(graph_data_list)-N+1))\n",
    "            assert selected_time_index >= 0, f\"select_graph_data: select_time_index {select_time_index} must be non-negative!\"\n",
    "            x.append(datetime.datetime.strptime(window[selected_time_index]['created_at'], \"%a %b %d %H:%M:%S %z %Y\"))\n",
    "            y.append(percentage(window))\n",
    "        while len(window) >= N:\n",
    "            window.pop(0)\n",
    "    return(x, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "graph_data_list = make_graph_data_list(label_data, time_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "x, y = select_graph_data(graph_data_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "ax.xaxis.set_major_formatter(mdates.DateFormatter(\"%m\"))\n",
    "ax.plot_date(x, y, fmt=\"-\")\n",
    "plt.ylim(0, 100)\n",
    "plt.xlabel(\"months in the years 2020 and 2021\")\n",
    "plt.ylabel(\"percentage support\")\n",
    "plt.title(f\"support for {TOPIC} (annotated data)\")\n",
    "plt.savefig(f\"annotated-{TOPIC}.png\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. Make mood graph"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_x_y(dates):\n",
    "    data = {}\n",
    "    for date in dates:\n",
    "        month = date.month\n",
    "        if month not in data:\n",
    "            data[month] = 0\n",
    "        data[month] += 1\n",
    "    return [x for x in sorted(data.keys())], [data[x] for x in sorted(data.keys())]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "relevant_data = { id_str: time_data.loc[id_str]['created_at'] for id_str in time_data.index if label_data[int(id_str)] == 'RELEVANT' }\n",
    "irrelevant_data = { id_str: time_data.loc[id_str]['created_at'] for id_str in time_data.index if label_data[int(id_str)] == 'IRRELEVANT' }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "dates_relevant = [datetime.datetime.strptime(relevant_data[id_str], \"%a %b %d %H:%M:%S %z %Y\") for id_str in relevant_data]\n",
    "dates_irrelevant = [datetime.datetime.strptime(irrelevant_data[id_str], \"%a %b %d %H:%M:%S %z %Y\") for id_str in irrelevant_data]\n",
    "x_relevant, y_relevant = make_x_y(dates_relevant)\n",
    "x_irrelevant, y_irrelevant = make_x_y(dates_irrelevant)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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": [
    "font = {'family' : 'DejaVu Sans',\n",
    "        'weight' : 'normal',\n",
    "        'size'   : 16}\n",
    "matplotlib.rc('font', **font)\n",
    "\n",
    "plt.figure(figsize=(12, 6))\n",
    "plt.bar(x_relevant, y_relevant, label=f\"expressing {TOPIC_EN} ({len(dates_relevant)})\")\n",
    "plt.bar(x_irrelevant, y_irrelevant, bottom=y_relevant, label=f\"other ({len(dates_irrelevant)})\")\n",
    "for i in range(0, len(y_relevant)):\n",
    "    plt.text(x_relevant[i] - 0.32, y_relevant[i] + 2, f\"{int(100*y_relevant[i]/(y_relevant[i] + y_irrelevant[i]))}%\")\n",
    "plt.ylim([0, 110])\n",
    "plt.xticks(range(1,13), [datetime.datetime(2020,m,1).strftime(\"%b\") for m in range(1, 13)])\n",
    "plt.xlabel(\"months in the year 2020\")\n",
    "plt.ylabel(f\"frequency {TOPIC_EN} tweets\")\n",
    "plt.title(f\"{len(label_data)} annotated {TOPIC_EN} tweets of 2020\")\n",
    "plt.legend(loc=\"upper left\")\n",
    "plt.savefig(f\"annotated_freqs_{TOPIC_EN}.png\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "month_counts_2020 = [ 15418685, 14851673, 21180942, 18715900, 18044679, 25944877,\n",
    "                      21840960, 20314042, 20118797, 21987100, 19393478, 20249944 ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_month_data(topic, year):\n",
    "    hours_matched_list = pd.read_csv(f\"csv/hours_matched_{year}_{topic}.csv\", index_col=None).to_dict(orient=\"records\")\n",
    "    hours_matched = {}\n",
    "    month_data = {}\n",
    "    for row in hours_matched_list:\n",
    "        month = row[\"hour\"][4:6]\n",
    "        if month not in month_data:\n",
    "            month_data[month] = row[\"matched\"]\n",
    "        else:\n",
    "            month_data[month] += row[\"matched\"]\n",
    "    return month_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "month_data = get_month_data(TOPIC, 2020)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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": [
    "y_relative = [ (list(month_data.values())[i] * y_relevant[i]/(y_relevant[i] + y_irrelevant[i])) / month_counts_2020[i] \n",
    "               for i in range(0, len(y_relevant)) ]\n",
    "\n",
    "plt.figure(figsize=(12, 6))\n",
    "plt.bar(x_relevant, y_relative, label=f\"expressing {TOPIC_EN}\")\n",
    "plt.xticks(range(1,13), [datetime.datetime(2020,m,1).strftime(\"%b\") for m in range(1, 13)])\n",
    "plt.yticks([])\n",
    "plt.xlabel(\"months in the year 2020\")\n",
    "plt.title(f\"{TOPIC_EN} based on annotated tweets of 2020\")\n",
    "plt.savefig(f\"{TOPIC_EN}_annotated_2020.png\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "python37",
   "language": "python",
   "name": "python37"
  },
  "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
}