twitter_mining-v2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import sys\n",
    "from tweepy import API\n",
    "from tweepy import OAuthHandler\n",
    "from tweepy import Cursor\n",
    "import json\n",
    "from collections import Counter\n",
    "import math\n",
    "import time\n",
    "import pandas as pd\n",
    "import csv\n",
    "import numpy as np\n",
    "from functools import reduce\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "sns.set()\n",
    "from sklearn.cluster import KMeans"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def getTwitterOAuth():\n",
    "    consumer_key = ''\n",
    "    consumer_secret = ''\n",
    "    access_token = ''\n",
    "    access_token_secret = ''\n",
    "    auth = OAuthHandler(consumer_key, consumer_secret)\n",
    "    auth.set_access_token(access_token,access_token_secret)\n",
    "    return auth\n",
    "\n",
    "def TwitterClient():\n",
    "    auth = getTwitterOAuth()\n",
    "    client = API(auth)\n",
    "    return client\n",
    "\n",
    "#global variables\n",
    "MAX_FRIENDS = 15000\n",
    "max_pages = math.ceil(MAX_FRIENDS/5000)\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def paginate(items,n):\n",
    "    for i in range(0, len(items),n):\n",
    "        yield items[i:i+n] "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "usrSet = set()\n",
    "usrSet2 = set()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# user friends\n",
    "usrset3 = set()\n",
    "\n",
    "def userFriends(user=\"AvokaAyebilla\"):\n",
    "    client = TwitterClient()\n",
    "    fname = \"{}_friends.jsonl\".format(user)\n",
    "    with open(fname, 'w') as f:\n",
    "        for friends in Cursor(client.friends_ids,screen_name=user).pages(max_pages):\n",
    "            for chunk in paginate(friends, 100):\n",
    "                users = client.lookup_users(user_ids=chunk)\n",
    "                for usr in users:\n",
    "                    if usr.location =='Ghana':\n",
    "                        usrset3.add(usr.screen_name)\n",
    "                        f.write(json.dumps(usr._json)+\"\\n\")\n",
    "            if len(friends) == 300:\n",
    "                print(\"More results available. Sleeping for 1 minute to avoid rate limit..\")\n",
    "                time.sleep(120)\n",
    "    #return usrSet2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def getUserTweets(user='AvokaAyebilla'):\n",
    "    client = TwitterClient()\n",
    "    fname = \"user_timeline_{}.jsonl\".format(user)\n",
    "    with open(fname, 'w') as f:\n",
    "        for page in Cursor(client.user_timeline, screen_name=user,count=200).pages(2):\n",
    "            for status in page:\n",
    "                f.write(json.dumps(status._json) + \"\\n\")\n",
    "\n",
    "\n",
    "def getHashtags(tweet):\n",
    "    entities = tweet.get('entities',{})\n",
    "    hashtags = entities.get('hashtags',[])\n",
    "    return [tag['text'].lower() for tag in hashtags]\n",
    "def tags(user='AvokaAyebilla', dictTags = {}): #dicTags is a dictionary containing the keywords as keys and \n",
    "                                                #the hashtags of interest as a list of values\n",
    "    stats = {'economy': 0, 'social':0, 'culture': 0,'health':0}\n",
    "    labels = ['economy','social','culture','health']                                                          #set all feature count to zeros.\n",
    "    fname = \"user_timeline_{}.jsonl\".format(user)\n",
    "    with open(fname, 'r') as f:\n",
    "        hashtags = [] #Counter()\n",
    "        for line in f:\n",
    "            tweet = json.loads(line)\n",
    "            tags_in_tweet = getHashtags(tweet)\n",
    "            if len(tags_in_tweet)>0:\n",
    "                hashtags.append(tags_in_tweet)\n",
    "        for tag in hashtags:\n",
    "            for tt in tag:\n",
    "                for key in labels:\n",
    "                    if tt in dictTags[key]:\n",
    "                        stats[key] =stats[key] + 1 \n",
    "    #print(hashtags)\n",
    "    return {user: list(stats.values())}\n",
    "                    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "Following are list of some identified top Ghanaians in different industries \n",
    "identified. Their twitter handles were obtained by searching their names on twitter\n",
    "We then get their friends and add it to form the top 1000 Ghanaians on twitter.\n",
    "Repeaton was disallowed by using the python set.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "uList = []\n",
    "htags = {'economy': ['yearofreturn','ghanaianbusiness','localbusiness',\n",
    "                     'ghanabeyondaid','2020budgetreview',\n",
    "                     'sdgs','beyondthelockdown','farmersdayghana'], \n",
    "         'social':['ghtourism','ghweddings','imwithher','tadigirls',\n",
    "                   'justiceforgeorgefloyd','girlchildeducation',\n",
    "                   'lutterodt','hushpuppi','ugandavsghana'], \n",
    "         'culture': ['independenceday','ghanamostbeautiful','ghana',\n",
    "                     'saynotocorruption', 'goodmorningremix','gmb2020',\n",
    "                     'nsmq','registertovote2020', 'mokobe'],\n",
    "         'health':['stayhome','socialdistancing','spreadcalmnotfear',\n",
    "                   'ghanacovid19','wearyourmask','covid19','ghhealth',\n",
    "                   'globalhealth','staysafe']}\n",
    "\n",
    "with open('Top_ghanaians_usrs.csv', 'rt') as f:\n",
    "        f.readline()\n",
    "        csv_reader = csv.reader(f)\n",
    "        for line in csv_reader:\n",
    "            if os.path.exists(\"user_timeline_{}.jsonl\".format(line[0])):\n",
    "                try:\n",
    "                    uList.append(tags(line[0], htags))\n",
    "                except Exception as e:\n",
    "                    print( \"Error while getting hashtags for  {}\".format(line[0]))\n",
    "                    print(e)\n",
    "                \n",
    "\n",
    "'''Output from calling the tags function is a dictionary with the user specified as the key and the \n",
    "four parameters as the a list value. In order to call this function manytimes for the many users we have,\n",
    "use a list or a dictionary to record every return of the function. In this case, a list is used. Manipulate this \n",
    "list of dictionary to get the output into a list for easy conversion to a pandas dataframe. You can do it\n",
    "differently but here how I did it with the functions below. note that, this written with the assumption that\n",
    "the data exist on your disk. If you don't have the data, you will first run the getUserTweets() method to \n",
    "download the timelines of the users you need. \n",
    "'''\n",
    "#prepare a data frame\n",
    "def get_dictkeys(listdic):\n",
    "    return list(listdic.keys())\n",
    "\n",
    "#returns the values of the dictionary as a list\n",
    "def get_dictvals(listdic):\n",
    "     return list(listdic.values())\n",
    "\n",
    "def dataframe():\n",
    "    keysList = []\n",
    "    valList = []\n",
    "    data = []\n",
    "    fkeyList = []\n",
    "    for item in uList:\n",
    "        keysList.append(get_dictkeys(item))\n",
    "        valList.append(get_dictvals(item))\n",
    "    for it in valList:\n",
    "        data.append(it[0])\n",
    "    for n in keysList:\n",
    "        fkeyList.append(n[0])  \n",
    "    df = pd.DataFrame(data, columns = [\"economy\", \"social\",\"culture\",\"health\"])\n",
    "    df.insert(0,'user',fkeyList)\n",
    "    return df\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "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>user</th>\n",
       "      <th>economy</th>\n",
       "      <th>social</th>\n",
       "      <th>culture</th>\n",
       "      <th>health</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>sambahflex</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>CoOwusu</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>AshesiIX</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>askrashida</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Ghanasoccernet</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>TransformingGh</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>29</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>KelOwusu</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>fellowblackman</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>kkdbod</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>talentshype</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             user  economy  social  culture  health\n",
       "0      sambahflex        0       0        1       0\n",
       "1         CoOwusu        0       0        0       0\n",
       "2        AshesiIX        0       0        0       0\n",
       "3      askrashida        0       0        0       0\n",
       "4  Ghanasoccernet        0       0        0       0\n",
       "5  TransformingGh        0       0       29       0\n",
       "6        KelOwusu        0       0        8       6\n",
       "7  fellowblackman        0       0        2       0\n",
       "8          kkdbod        1       0        2       1\n",
       "9     talentshype        0       0        0       0"
      ]
     },
     "execution_count": 216,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dff = dataframe()\n",
    "dff.to_csv('ghanadata.csv', sep=',', index =False, encoding='utf-8')\n",
    "dff.head(10)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = dff.iloc[:,1:5]\n",
    "kmeans = KMeans(4)\n",
    "#kmeans.fit(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 0, ..., 0, 0, 0])"
      ]
     },
     "execution_count": 207,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "identified_cluster = kmeans.fit_predict(X)\n",
    "identified_cluster"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "data_wt_clusters = dff.copy()\n",
    "data_wt_clusters['clusters'] = identified_cluster\n",
    "fig =plt.figure()\n",
    "#ax = Axes3D(fig, rect=[0, 0, .95, 1], elev=48, azim=134)\n",
    "#ax = fig.add_subplot(111,projection='3d')\n",
    "ax = Axes3D(fig, rect=[1, 1, 1, 1], elev=48, azim=200)\n",
    "\n",
    "x = data_wt_clusters['economy']\n",
    "y = data_wt_clusters['social']\n",
    "z = data_wt_clusters['culture']\n",
    "c = data_wt_clusters['health']\n",
    "\n",
    "img = ax.scatter(x,y,z, c=c, cmap=plt.hot())\n",
    "fig.colorbar(img)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Processing the data required many discrete step that required hard-coding at some points. The functions below were used to process the lists of friends to store them into a csv file as one. Once we have the complete csv file of uses, we really don't these functions anymore. I have left just in case I want to extend my input then I may just come back to them again. The major thing is to read and understand the key helper functions above and know what they return. You then decide how you want to manipulate the results returned to you. You also want to understand the input and consider the time it takes to fetch the data you want from twitter. with this in mind, you can run this code step by step to get the desired results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "handset1 = set()\n",
    "handset2 = set()\n",
    "handset3 = set()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 227,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "handles = []\n",
    "def readfile():\n",
    "    with open('Top_ghanaians.csv', 'rt') as f:\n",
    "        f.readline()\n",
    "        csv_reader = csv.reader(f)\n",
    "        for line in csv_reader:\n",
    "            handles.append(line)         \n",
    "readfile()           "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 228,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in handles:\n",
    "    for j in i:\n",
    "        handset1.add(j)\n",
    "#handset2\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "metadata": {},
   "outputs": [],
   "source": [
    "handset4 = set()\n",
    "userList = useList + list3 + list2\n",
    "for i in userList:\n",
    "    handset4.add(i)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {},
   "outputs": [],
   "source": [
    "usrhandles = set.union(handset1, handset2, handset3,handset4)\n",
    "df = pd.DataFrame(usrhandles, columns = ['user_handles'])\n",
    "df.to_csv('Top_ghanaians_usrs.csv', sep=',', index =False, encoding='utf-8')"
   ]
  }
 ],
 "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}