BTC_sentiment_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "4e4253c8",
   "metadata": {},
   "source": [
    "# BTC sentiment analysis\n",
    "\n",
    "With this script, I will download and analyze 100.000 Bitcoin (BTC) tweets from the last two weeks and perform sentiment analysis to gather market intelligence. The aim of the analysis is to understand what are people’s opinions about BTC tweets. \n",
    "\n",
    "Sentiment analysis consists of converting a text into a score that estimates its sentiment (e.g., demand for a given product) and can be deemed as a classification task (e.e., to extract subjectivity and sentiment polarity from text data).\n",
    "\n",
    "Web scraping is done using snscrape (https://github.com/JustAnotherArchivist/snscrape). \\\n",
    "Filtering for English language is done using langdetect (https://github.com/Mimino666/langdetect). \\\n",
    "Sentiment analysis is done using spacytextblob (https://github.com/SamEdwardes/spacytextblob).\n",
    "\n",
    "This script will perform the following:\n",
    "1. Download data from Twitter (using snscrape)\n",
    "2. Preprocess the data\n",
    "3. Perform sentiment analysis (using spacytextblob)\n",
    "4. Analyze & visualize the results (using nltk for preproc.)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0c05d494",
   "metadata": {},
   "source": [
    "### 1. Download data from Twitter\n",
    "\n",
    "To download data from Twitter without using its metered API, hence without any limit on the volume of data I wish to scrape.\n",
    "\n",
    "After loading the libraries, I will declare the search parameters. Specifically, I will be scraping 100.000 Tweets from Nov 1 2022 to yesterday (Nov 15 2022)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "da72cca5",
   "metadata": {},
   "outputs": [],
   "source": [
    "import snscrape.modules.twitter as sntwitter\n",
    "import progressbar\n",
    "from time import sleep\n",
    "from datetime import datetime\n",
    "import os\n",
    "import string\n",
    "import re\n",
    "import statistics\n",
    "from collections import Counter\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# for removing non-Eng tweets:\n",
    "from langdetect import detect\n",
    "from langdetect import detect_langs\n",
    "\n",
    "# for performing sentiment analysis:\n",
    "import spacy\n",
    "from spacytextblob.spacytextblob import SpacyTextBlob\n",
    "\n",
    "# for lemmatization:\n",
    "import nltk\n",
    "from nltk.tokenize import RegexpTokenizer\n",
    "from nltk.stem import WordNetLemmatizer,PorterStemmer\n",
    "from nltk.corpus import stopwords"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "3add83ca",
   "metadata": {},
   "outputs": [],
   "source": [
    "BTC_dict = {'bitcoin': ['bitcoin since:2022-11-01 until:2022-11-15', 100000]}\n",
    "output_name = 'BTC_100k_tweets_until_2022-11-15.csv'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "29c75034",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[                                                                        ]   0%\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "bitcoin %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[========================================================================] 100%\n"
     ]
    }
   ],
   "source": [
    "today = datetime.today().strftime('%Y%m%d')[2:]+'_'\n",
    "for index, BTC_name in enumerate(BTC_dict):\n",
    "    print(BTC_name, '%')\n",
    "    tweets_list1 = []\n",
    "    bar = progressbar.ProgressBar(maxval=BTC_dict[BTC_name][1]+2, widgets=[progressbar.Bar('=', '[', ']'), ' ', progressbar.Percentage()])\n",
    "    bar.start()\n",
    "    for i,tweet in enumerate(sntwitter.TwitterSearchScraper(f'{BTC_dict[BTC_name][0]}').get_items()): #declare a username\n",
    "        bar.update(i+1)\n",
    "        if i>BTC_dict[BTC_name][1]: #number of tweets you want to scrape\n",
    "            break\n",
    "        #print(movie_name, i, tweet)\n",
    "        tweets_list1.append([tweet.date, tweet.id, tweet.content, tweet.user.username]) #declare the attributes to be returned\n",
    "    tweets_df1 = pd.DataFrame(tweets_list1, columns=['Datetime', 'Tweet Id', 'Text', 'Username'])\n",
    "\n",
    "    tweets_df1[['Datetime', 'Text']].to_csv(output_name)\n",
    "    bar.finish()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6d11f65a",
   "metadata": {},
   "source": [
    "### 2. Preprocess the data\n",
    "\n",
    "Let's take a look at the output file."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "44cca1ec",
   "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>Text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>99991</th>\n",
       "      <td>Plot twist: \\n\\nCIA, FBI &amp;amp; SEC worked toge...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99992</th>\n",
       "      <td>@CryptoDonAlt Time to become a #bitcoin maxi.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99993</th>\n",
       "      <td>1 CORE = 1 Ethereum\\n*🏆CORE Satoshi Mining🏆* h...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99994</th>\n",
       "      <td>Bitcoin, Ethereum, Chain, and EOS Daily Price ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99995</th>\n",
       "      <td>@ishtarmol @joseantonetti Jajajajajajajaja si,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99996</th>\n",
       "      <td>#BinanceCoin price update: $BNB\\n\\n#BNB $278.3...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99997</th>\n",
       "      <td>@RealJamesWoods What about #bitcoin though Jam...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99998</th>\n",
       "      <td>US Treasury Secretary and Indian Finance Minis...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99999</th>\n",
       "      <td>#bitcoin Satoshi whitepaper vision is pure and...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100000</th>\n",
       "      <td>@EntropySloth @woed051 @RayVanlandingh1 @Alice...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                     Text\n",
       "99991   Plot twist: \\n\\nCIA, FBI &amp; SEC worked toge...\n",
       "99992       @CryptoDonAlt Time to become a #bitcoin maxi.\n",
       "99993   1 CORE = 1 Ethereum\\n*🏆CORE Satoshi Mining🏆* h...\n",
       "99994   Bitcoin, Ethereum, Chain, and EOS Daily Price ...\n",
       "99995   @ishtarmol @joseantonetti Jajajajajajajaja si,...\n",
       "99996   #BinanceCoin price update: $BNB\\n\\n#BNB $278.3...\n",
       "99997   @RealJamesWoods What about #bitcoin though Jam...\n",
       "99998   US Treasury Secretary and Indian Finance Minis...\n",
       "99999   #bitcoin Satoshi whitepaper vision is pure and...\n",
       "100000  @EntropySloth @woed051 @RayVanlandingh1 @Alice..."
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv(output_name)[['Text']]\n",
    "df.tail(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "98411a6b",
   "metadata": {},
   "source": [
    "Let's get rid of all the @, #, and non-ascii characters, and links, and take a look again:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "279bf2b6",
   "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>Text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>99991</th>\n",
       "      <td>Plot twist CIA FBI amp SEC worked together to ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99992</th>\n",
       "      <td>Time to become a bitcoin maxi</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99993</th>\n",
       "      <td>1 CORE 1 Ethereum CORE Satoshi Mining of Mr Sa...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99994</th>\n",
       "      <td>Bitcoin Ethereum Chain and EOS Daily Price Ana...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99995</th>\n",
       "      <td>Jajajajajajajaja si ya veo FTX quebró el bitco...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99996</th>\n",
       "      <td>BinanceCoin price update BNB BNB 278 34 USD B...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99997</th>\n",
       "      <td>What about bitcoin though James Have you learn...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99998</th>\n",
       "      <td>US Treasury Secretary and Indian Finance Minis...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99999</th>\n",
       "      <td>bitcoin Satoshi whitepaper vision is pure and...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100000</th>\n",
       "      <td>Bitcoin isn t currency But otherwise this has ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                     Text\n",
       "99991   Plot twist CIA FBI amp SEC worked together to ...\n",
       "99992                      Time to become a bitcoin maxi \n",
       "99993   1 CORE 1 Ethereum CORE Satoshi Mining of Mr Sa...\n",
       "99994   Bitcoin Ethereum Chain and EOS Daily Price Ana...\n",
       "99995   Jajajajajajajaja si ya veo FTX quebró el bitco...\n",
       "99996    BinanceCoin price update BNB BNB 278 34 USD B...\n",
       "99997   What about bitcoin though James Have you learn...\n",
       "99998   US Treasury Secretary and Indian Finance Minis...\n",
       "99999    bitcoin Satoshi whitepaper vision is pure and...\n",
       "100000  Bitcoin isn t currency But otherwise this has ..."
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#get rid of links and hashtags\n",
    "df[\"Text\"] = df[\"Text\"].apply(lambda x : ' '.join([s for s in x.split(' ') if s.find('@') == -1 and s.find('www') == -1 and s.find('https') == -1]))\n",
    "\n",
    "#get rid of non-ascii characters\n",
    "df = df.replace(r'\\W+', ' ', regex=True)\n",
    "df.tail(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0b9edf59",
   "metadata": {},
   "source": [
    "The next preprocessing step is to remove empty and non-English tweets: training a neural netwoek to acomplish sentiment analysis is already hard with one language!\n",
    "\n",
    "Incidentally, I also remove cells that do not contain text but for some reason do not show up to be empty when running commands such as np.where(pd.isnull(df))\n",
    "\n",
    "First, we populate a vector describing whether the tweet is readable and in English (using the detect module of langdetect) (=1) or not (=0) and we then remove all the rows with index 0."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "d0639dd4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Row  183  has some issue: remove\n",
      "Row  796  has some issue: remove\n",
      "Row  1341  has some issue: remove\n",
      "Row  1480  has some issue: remove\n",
      "Row  1712  has some issue: remove\n",
      "Row  1749  has some issue: remove\n",
      "Row  2096  has some issue: remove\n",
      "Row  2133  has some issue: remove\n",
      "Row  2230  has some issue: remove\n",
      "Row  2321  has some issue: remove\n",
      "Row  2400  has some issue: remove\n",
      "Row  2456  has some issue: remove\n",
      "Row  2517  has some issue: remove\n",
      "Row  2563  has some issue: remove\n",
      "Row  3085  has some issue: remove\n",
      "Row  3099  has some issue: remove\n",
      "Row  3102  has some issue: remove\n",
      "Row  3178  has some issue: remove\n",
      "Row  3211  has some issue: remove\n",
      "Row  3304  has some issue: remove\n",
      "Row  3586  has some issue: remove\n",
      "Row  4358  has some issue: remove\n",
      "Row  4531  has some issue: remove\n",
      "Row  4842  has some issue: remove\n",
      "Row  4888  has some issue: remove\n",
      "Row  5016  has some issue: remove\n",
      "Row  5237  has some issue: remove\n",
      "Row  5360  has some issue: remove\n",
      "Row  5684  has some issue: remove\n",
      "Row  5854  has some issue: remove\n",
      "Row  5870  has some issue: remove\n",
      "Row  6548  has some issue: remove\n",
      "Row  6825  has some issue: remove\n",
      "Row  7141  has some issue: remove\n",
      "Row  7713  has some issue: remove\n",
      "Row  7802  has some issue: remove\n",
      "Row  8094  has some issue: remove\n",
      "Row  8168  has some issue: remove\n",
      "Row  8355  has some issue: remove\n",
      "Row  9674  has some issue: remove\n",
      "Row  9695  has some issue: remove\n",
      "Row  9802  has some issue: remove\n",
      "Row  9812  has some issue: remove\n",
      "Row  10184  has some issue: remove\n",
      "Row  10193  has some issue: remove\n",
      "Row  10335  has some issue: remove\n",
      "Row  10375  has some issue: remove\n",
      "Row  10459  has some issue: remove\n",
      "Row  10787  has some issue: remove\n",
      "Row  10857  has some issue: remove\n",
      "Row  10981  has some issue: remove\n",
      "Row  11027  has some issue: remove\n",
      "Row  11292  has some issue: remove\n",
      "Row  11538  has some issue: remove\n",
      "Row  11699  has some issue: remove\n",
      "Row  11854  has some issue: remove\n",
      "Row  11917  has some issue: remove\n",
      "Row  12135  has some issue: remove\n",
      "Row  12332  has some issue: remove\n",
      "Row  12391  has some issue: remove\n",
      "Row  12545  has some issue: remove\n",
      "Row  13161  has some issue: remove\n",
      "Row  13266  has some issue: remove\n",
      "Row  13327  has some issue: remove\n",
      "Row  13621  has some issue: remove\n",
      "Row  13863  has some issue: remove\n",
      "Row  14126  has some issue: remove\n",
      "Row  14267  has some issue: remove\n",
      "Row  14336  has some issue: remove\n",
      "Row  14454  has some issue: remove\n",
      "Row  14627  has some issue: remove\n",
      "Row  14914  has some issue: remove\n",
      "Row  14992  has some issue: remove\n",
      "Row  15009  has some issue: remove\n",
      "Row  15166  has some issue: remove\n",
      "Row  15253  has some issue: remove\n",
      "Row  15478  has some issue: remove\n",
      "Row  15527  has some issue: remove\n",
      "Row  15550  has some issue: remove\n",
      "Row  15702  has some issue: remove\n",
      "Row  15756  has some issue: remove\n",
      "Row  15852  has some issue: remove\n",
      "Row  15900  has some issue: remove\n",
      "Row  15904  has some issue: remove\n",
      "Row  15950  has some issue: remove\n",
      "Row  16014  has some issue: remove\n",
      "Row  16226  has some issue: remove\n",
      "Row  16800  has some issue: remove\n",
      "Row  17469  has some issue: remove\n",
      "Row  17583  has some issue: remove\n",
      "Row  17807  has some issue: remove\n",
      "Row  18063  has some issue: remove\n",
      "Row  18490  has some issue: remove\n",
      "Row  18584  has some issue: remove\n",
      "Row  18594  has some issue: remove\n",
      "Row  18917  has some issue: remove\n",
      "Row  19067  has some issue: remove\n",
      "Row  19234  has some issue: remove\n",
      "Row  19517  has some issue: remove\n",
      "Row  19834  has some issue: remove\n",
      "Row  19925  has some issue: remove\n",
      "Row  20472  has some issue: remove\n",
      "Row  20476  has some issue: remove\n",
      "Row  20666  has some issue: remove\n",
      "Row  20826  has some issue: remove\n",
      "Row  20985  has some issue: remove\n",
      "Row  21151  has some issue: remove\n",
      "Row  21610  has some issue: remove\n",
      "Row  21706  has some issue: remove\n",
      "Row  21794  has some issue: remove\n",
      "Row  22320  has some issue: remove\n",
      "Row  22332  has some issue: remove\n",
      "Row  23070  has some issue: remove\n",
      "Row  23135  has some issue: remove\n",
      "Row  23138  has some issue: remove\n",
      "Row  23436  has some issue: remove\n",
      "Row  23740  has some issue: remove\n",
      "Row  24793  has some issue: remove\n",
      "Row  25090  has some issue: remove\n",
      "Row  25115  has some issue: remove\n",
      "Row  25630  has some issue: remove\n",
      "Row  26000  has some issue: remove\n",
      "Row  26458  has some issue: remove\n",
      "Row  26807  has some issue: remove\n",
      "Row  26963  has some issue: remove\n",
      "Row  27421  has some issue: remove\n",
      "Row  27559  has some issue: remove\n",
      "Row  27826  has some issue: remove\n",
      "Row  27947  has some issue: remove\n",
      "Row  28034  has some issue: remove\n",
      "Row  28158  has some issue: remove\n",
      "Row  28214  has some issue: remove\n",
      "Row  28451  has some issue: remove\n",
      "Row  28468  has some issue: remove\n",
      "Row  29088  has some issue: remove\n",
      "Row  29275  has some issue: remove\n",
      "Row  29352  has some issue: remove\n",
      "Row  29727  has some issue: remove\n",
      "Row  29895  has some issue: remove\n",
      "Row  30118  has some issue: remove\n",
      "Row  30236  has some issue: remove\n",
      "Row  30505  has some issue: remove\n",
      "Row  30568  has some issue: remove\n",
      "Row  30694  has some issue: remove\n",
      "Row  30745  has some issue: remove\n",
      "Row  30855  has some issue: remove\n",
      "Row  30876  has some issue: remove\n",
      "Row  31039  has some issue: remove\n",
      "Row  31075  has some issue: remove\n",
      "Row  31101  has some issue: remove\n",
      "Row  31171  has some issue: remove\n",
      "Row  31328  has some issue: remove\n",
      "Row  31580  has some issue: remove\n",
      "Row  31689  has some issue: remove\n",
      "Row  31737  has some issue: remove\n",
      "Row  31787  has some issue: remove\n",
      "Row  31819  has some issue: remove\n",
      "Row  32035  has some issue: remove\n",
      "Row  32205  has some issue: remove\n",
      "Row  32265  has some issue: remove\n",
      "Row  32330  has some issue: remove\n",
      "Row  32528  has some issue: remove\n",
      "Row  32532  has some issue: remove\n",
      "Row  32655  has some issue: remove\n",
      "Row  32787  has some issue: remove\n",
      "Row  33001  has some issue: remove\n",
      "Row  33267  has some issue: remove\n",
      "Row  33287  has some issue: remove\n",
      "Row  33290  has some issue: remove\n",
      "Row  33385  has some issue: remove\n",
      "Row  33419  has some issue: remove\n",
      "Row  33458  has some issue: remove\n",
      "Row  33470  has some issue: remove\n",
      "Row  33516  has some issue: remove\n",
      "Row  33586  has some issue: remove\n",
      "Row  33661  has some issue: remove\n",
      "Row  33678  has some issue: remove\n",
      "Row  33682  has some issue: remove\n",
      "Row  33885  has some issue: remove\n",
      "Row  34815  has some issue: remove\n",
      "Row  35086  has some issue: remove\n",
      "Row  35786  has some issue: remove\n",
      "Row  36587  has some issue: remove\n",
      "Row  36644  has some issue: remove\n",
      "Row  37817  has some issue: remove\n",
      "Row  38029  has some issue: remove\n",
      "Row  38350  has some issue: remove\n",
      "Row  38402  has some issue: remove\n",
      "Row  38733  has some issue: remove\n",
      "Row  39127  has some issue: remove\n",
      "Row  39339  has some issue: remove\n",
      "Row  39378  has some issue: remove\n",
      "Row  40759  has some issue: remove\n",
      "Row  40810  has some issue: remove\n",
      "Row  40842  has some issue: remove\n",
      "Row  41044  has some issue: remove\n",
      "Row  41458  has some issue: remove\n",
      "Row  41546  has some issue: remove\n",
      "Row  41579  has some issue: remove\n",
      "Row  41583  has some issue: remove\n",
      "Row  41589  has some issue: remove\n",
      "Row  41615  has some issue: remove\n",
      "Row  41686  has some issue: remove\n",
      "Row  41756  has some issue: remove\n",
      "Row  41825  has some issue: remove\n",
      "Row  41877  has some issue: remove\n",
      "Row  41889  has some issue: remove\n",
      "Row  41922  has some issue: remove\n",
      "Row  41942  has some issue: remove\n",
      "Row  42059  has some issue: remove\n",
      "Row  42189  has some issue: remove\n",
      "Row  42222  has some issue: remove\n",
      "Row  42408  has some issue: remove\n",
      "Row  42653  has some issue: remove\n",
      "Row  43260  has some issue: remove\n",
      "Row  43468  has some issue: remove\n",
      "Row  43616  has some issue: remove\n",
      "Row  43868  has some issue: remove\n",
      "Row  44075  has some issue: remove\n",
      "Row  44344  has some issue: remove\n",
      "Row  44359  has some issue: remove\n",
      "Row  44871  has some issue: remove\n",
      "Row  44881  has some issue: remove\n",
      "Row  45161  has some issue: remove\n",
      "Row  45164  has some issue: remove\n",
      "Row  45274  has some issue: remove\n",
      "Row  45802  has some issue: remove\n",
      "Row  46528  has some issue: remove\n",
      "Row  47126  has some issue: remove\n",
      "Row  47261  has some issue: remove\n",
      "Row  47264  has some issue: remove\n",
      "Row  48025  has some issue: remove\n",
      "Row  48591  has some issue: remove\n",
      "Row  50053  has some issue: remove\n",
      "Row  50074  has some issue: remove\n",
      "Row  50445  has some issue: remove\n",
      "Row  50463  has some issue: remove\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Row  50559  has some issue: remove\n",
      "Row  51543  has some issue: remove\n",
      "Row  51603  has some issue: remove\n",
      "Row  51688  has some issue: remove\n",
      "Row  51879  has some issue: remove\n",
      "Row  51991  has some issue: remove\n",
      "Row  52897  has some issue: remove\n",
      "Row  53909  has some issue: remove\n",
      "Row  54159  has some issue: remove\n",
      "Row  54589  has some issue: remove\n",
      "Row  54674  has some issue: remove\n",
      "Row  55605  has some issue: remove\n",
      "Row  55643  has some issue: remove\n",
      "Row  55649  has some issue: remove\n",
      "Row  55865  has some issue: remove\n",
      "Row  55926  has some issue: remove\n",
      "Row  55944  has some issue: remove\n",
      "Row  56353  has some issue: remove\n",
      "Row  56783  has some issue: remove\n",
      "Row  57327  has some issue: remove\n",
      "Row  57542  has some issue: remove\n",
      "Row  57558  has some issue: remove\n",
      "Row  57702  has some issue: remove\n",
      "Row  57930  has some issue: remove\n",
      "Row  58158  has some issue: remove\n",
      "Row  58247  has some issue: remove\n",
      "Row  58281  has some issue: remove\n",
      "Row  58671  has some issue: remove\n",
      "Row  58769  has some issue: remove\n",
      "Row  58838  has some issue: remove\n",
      "Row  58878  has some issue: remove\n",
      "Row  58923  has some issue: remove\n",
      "Row  58971  has some issue: remove\n",
      "Row  59058  has some issue: remove\n",
      "Row  59317  has some issue: remove\n",
      "Row  59460  has some issue: remove\n",
      "Row  59664  has some issue: remove\n",
      "Row  59917  has some issue: remove\n",
      "Row  59919  has some issue: remove\n",
      "Row  60058  has some issue: remove\n",
      "Row  60103  has some issue: remove\n",
      "Row  60259  has some issue: remove\n",
      "Row  60337  has some issue: remove\n",
      "Row  60451  has some issue: remove\n",
      "Row  60980  has some issue: remove\n",
      "Row  61460  has some issue: remove\n",
      "Row  61579  has some issue: remove\n",
      "Row  61647  has some issue: remove\n",
      "Row  61679  has some issue: remove\n",
      "Row  61757  has some issue: remove\n",
      "Row  62019  has some issue: remove\n",
      "Row  62514  has some issue: remove\n",
      "Row  63020  has some issue: remove\n",
      "Row  63038  has some issue: remove\n",
      "Row  63156  has some issue: remove\n",
      "Row  63239  has some issue: remove\n",
      "Row  63264  has some issue: remove\n",
      "Row  63279  has some issue: remove\n",
      "Row  63420  has some issue: remove\n",
      "Row  63436  has some issue: remove\n",
      "Row  63448  has some issue: remove\n",
      "Row  63488  has some issue: remove\n",
      "Row  63588  has some issue: remove\n",
      "Row  64022  has some issue: remove\n",
      "Row  64205  has some issue: remove\n",
      "Row  64605  has some issue: remove\n",
      "Row  64851  has some issue: remove\n",
      "Row  65117  has some issue: remove\n",
      "Row  65345  has some issue: remove\n",
      "Row  65693  has some issue: remove\n",
      "Row  66177  has some issue: remove\n",
      "Row  66825  has some issue: remove\n",
      "Row  66827  has some issue: remove\n",
      "Row  66853  has some issue: remove\n",
      "Row  66919  has some issue: remove\n",
      "Row  67942  has some issue: remove\n",
      "Row  68240  has some issue: remove\n",
      "Row  68302  has some issue: remove\n",
      "Row  68382  has some issue: remove\n",
      "Row  68400  has some issue: remove\n",
      "Row  68640  has some issue: remove\n",
      "Row  69145  has some issue: remove\n",
      "Row  69441  has some issue: remove\n",
      "Row  69603  has some issue: remove\n",
      "Row  69637  has some issue: remove\n",
      "Row  69744  has some issue: remove\n",
      "Row  69767  has some issue: remove\n",
      "Row  69866  has some issue: remove\n",
      "Row  70645  has some issue: remove\n",
      "Row  70846  has some issue: remove\n",
      "Row  71093  has some issue: remove\n",
      "Row  71290  has some issue: remove\n",
      "Row  72086  has some issue: remove\n",
      "Row  72124  has some issue: remove\n",
      "Row  72185  has some issue: remove\n",
      "Row  72233  has some issue: remove\n",
      "Row  72823  has some issue: remove\n",
      "Row  72940  has some issue: remove\n",
      "Row  73329  has some issue: remove\n",
      "Row  73333  has some issue: remove\n",
      "Row  73470  has some issue: remove\n",
      "Row  73691  has some issue: remove\n",
      "Row  73838  has some issue: remove\n",
      "Row  73906  has some issue: remove\n",
      "Row  73923  has some issue: remove\n",
      "Row  74092  has some issue: remove\n",
      "Row  74191  has some issue: remove\n",
      "Row  74221  has some issue: remove\n",
      "Row  74235  has some issue: remove\n",
      "Row  74283  has some issue: remove\n",
      "Row  75028  has some issue: remove\n",
      "Row  75234  has some issue: remove\n",
      "Row  75882  has some issue: remove\n",
      "Row  75892  has some issue: remove\n",
      "Row  75955  has some issue: remove\n",
      "Row  75983  has some issue: remove\n",
      "Row  76287  has some issue: remove\n",
      "Row  76431  has some issue: remove\n",
      "Row  76432  has some issue: remove\n",
      "Row  76993  has some issue: remove\n",
      "Row  77178  has some issue: remove\n",
      "Row  77478  has some issue: remove\n",
      "Row  77792  has some issue: remove\n",
      "Row  78229  has some issue: remove\n",
      "Row  78716  has some issue: remove\n",
      "Row  78772  has some issue: remove\n",
      "Row  79367  has some issue: remove\n",
      "Row  79665  has some issue: remove\n",
      "Row  79715  has some issue: remove\n",
      "Row  80657  has some issue: remove\n",
      "Row  80859  has some issue: remove\n",
      "Row  80873  has some issue: remove\n",
      "Row  81402  has some issue: remove\n",
      "Row  81643  has some issue: remove\n",
      "Row  81846  has some issue: remove\n",
      "Row  82036  has some issue: remove\n",
      "Row  82219  has some issue: remove\n",
      "Row  82259  has some issue: remove\n",
      "Row  82276  has some issue: remove\n",
      "Row  82307  has some issue: remove\n",
      "Row  82833  has some issue: remove\n",
      "Row  83119  has some issue: remove\n",
      "Row  83208  has some issue: remove\n",
      "Row  83434  has some issue: remove\n",
      "Row  83692  has some issue: remove\n",
      "Row  83764  has some issue: remove\n",
      "Row  84137  has some issue: remove\n",
      "Row  84296  has some issue: remove\n",
      "Row  84355  has some issue: remove\n",
      "Row  84460  has some issue: remove\n",
      "Row  84873  has some issue: remove\n",
      "Row  85141  has some issue: remove\n",
      "Row  85625  has some issue: remove\n",
      "Row  85821  has some issue: remove\n",
      "Row  85822  has some issue: remove\n",
      "Row  85877  has some issue: remove\n",
      "Row  86501  has some issue: remove\n",
      "Row  86694  has some issue: remove\n",
      "Row  87518  has some issue: remove\n",
      "Row  88346  has some issue: remove\n",
      "Row  88463  has some issue: remove\n",
      "Row  88834  has some issue: remove\n",
      "Row  89048  has some issue: remove\n",
      "Row  89623  has some issue: remove\n",
      "Row  90051  has some issue: remove\n",
      "Row  91432  has some issue: remove\n",
      "Row  91765  has some issue: remove\n",
      "Row  91795  has some issue: remove\n",
      "Row  92600  has some issue: remove\n",
      "Row  92630  has some issue: remove\n",
      "Row  92841  has some issue: remove\n",
      "Row  93056  has some issue: remove\n",
      "Row  93249  has some issue: remove\n",
      "Row  93264  has some issue: remove\n",
      "Row  93598  has some issue: remove\n",
      "Row  93698  has some issue: remove\n",
      "Row  93700  has some issue: remove\n",
      "Row  93964  has some issue: remove\n",
      "Row  94261  has some issue: remove\n",
      "Row  94511  has some issue: remove\n",
      "Row  94642  has some issue: remove\n",
      "Row  95466  has some issue: remove\n",
      "Row  96109  has some issue: remove\n",
      "Row  96319  has some issue: remove\n",
      "Row  96394  has some issue: remove\n",
      "Row  96442  has some issue: remove\n",
      "Row  96792  has some issue: remove\n",
      "Row  96986  has some issue: remove\n",
      "Row  97068  has some issue: remove\n",
      "Row  97073  has some issue: remove\n",
      "Row  97393  has some issue: remove\n",
      "Row  97852  has some issue: remove\n",
      "Row  98135  has some issue: remove\n",
      "Row  98371  has some issue: remove\n",
      "Row  98557  has some issue: remove\n",
      "Row  98851  has some issue: remove\n",
      "Row  98947  has some issue: remove\n",
      "Row  99097  has some issue: remove\n",
      "Row  99195  has some issue: remove\n",
      "Row  99240  has some issue: remove\n",
      "Row  99333  has some issue: remove\n",
      "Row  99346  has some issue: remove\n",
      "Row  99348  has some issue: remove\n",
      "Row  99388  has some issue: remove\n",
      "Row  99429  has some issue: remove\n",
      "Row  99639  has some issue: remove\n",
      "Row  99783  has some issue: remove\n",
      "Row  99984  has some issue: remove\n"
     ]
    }
   ],
   "source": [
    "isEng = [0]*len(df)\n",
    "\n",
    "for i in range(len(df)):\n",
    "    try:\n",
    "        isEngTxt = detect(df.loc[i, \"Text\"])\n",
    "        if isEngTxt == 'en':\n",
    "            isEng[i] = 1\n",
    "    except:\n",
    "        print(f\"Row \", str(i), \" has some issue: remove\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "39329a9d",
   "metadata": {},
   "outputs": [],
   "source": [
    "indices = [\n",
    "    index for index, item in enumerate(isEng)\n",
    "    if item != 0\n",
    "]\n",
    "df = df[df.index.isin(indices)]\n",
    "df.tail(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "215d25bb",
   "metadata": {},
   "source": [
    "### 3. Perform sentiment analysis\n",
    "\n",
    "I am now going to apply a sentiment analysis to the cleaned data.\n",
    "\n",
    "To install spacytextblob and download corpora: \\\n",
    "!pip install spacytexblob==3.0.1 \\\n",
    "!pip install spacy==3.2.1 \\\n",
    "!python -m textblob.download_corpora \\\n",
    "!python -m spacy download en_core_web_sm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "9a9b6008",
   "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>Text</th>\n",
       "      <th>sentiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>After Bitcoin s Worst Week in Five Months Here...</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>After Bitcoin s Worst Week in Five Months Here...</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>After Bitcoin s Worst Week in Five Months Here...</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>After Bitcoin s Worst Week in Five Months Here...</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>After Bitcoin s Worst Week in Five Months Here...</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77633</th>\n",
       "      <td>Ted is the best dev in the space 100xGems cryp...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77634</th>\n",
       "      <td>Rolled tacos Feedgoat2000 found bitcoin in a U...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77635</th>\n",
       "      <td>psychojellybean found bitcoin in a User vault ...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77636</th>\n",
       "      <td>Buying Bitcoin today could be your best decisi...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77637</th>\n",
       "      <td>PonchoValdez found bitcoin in a User vault at ...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>77638 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    Text  sentiment\n",
       "0      After Bitcoin s Worst Week in Five Months Here...       -1.0\n",
       "1      After Bitcoin s Worst Week in Five Months Here...       -1.0\n",
       "2      After Bitcoin s Worst Week in Five Months Here...       -1.0\n",
       "3      After Bitcoin s Worst Week in Five Months Here...       -1.0\n",
       "4      After Bitcoin s Worst Week in Five Months Here...       -1.0\n",
       "...                                                  ...        ...\n",
       "77633  Ted is the best dev in the space 100xGems cryp...        1.0\n",
       "77634  Rolled tacos Feedgoat2000 found bitcoin in a U...        1.0\n",
       "77635  psychojellybean found bitcoin in a User vault ...        1.0\n",
       "77636  Buying Bitcoin today could be your best decisi...        1.0\n",
       "77637  PonchoValdez found bitcoin in a User vault at ...        1.0\n",
       "\n",
       "[77638 rows x 2 columns]"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nlp = spacy.load('en_core_web_sm')\n",
    "nlp.add_pipe(\"spacytextblob\")\n",
    "\n",
    "df['sentiment'] = df['Text'].apply(lambda x : nlp(x)._.polarity)\n",
    "df_sentiment = df.sort_values('sentiment').reset_index(drop=True)\n",
    "df_sentiment"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "38125e08",
   "metadata": {},
   "source": [
    "### 4. Analyze results\n",
    "\n",
    "Before analyzing the content of the tweets, we are first going to preprocess our data even more. Specifically, I am going to:\n",
    "\n",
    "- Lemmatize each word\n",
    "- Delete extra characters\n",
    "- Remove stop words\n",
    "\n",
    "Lemmatization will be carried out using the WordNetLemmatizer. The function also makes it possible to do stemming in place of (or addition to) lemmatization. I chose lemmatization in place of stemming as the vocab space is relatively large compared to the length of the document.\n",
    "\n",
    "Before running the function, download: \\\n",
    "nltk.download('wordnet') \\\n",
    "nltk.download('stopwords') \\\n",
    "nltk.download('omw-1.4')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "e24d2c3b",
   "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>Text</th>\n",
       "      <th>sentiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>bitcoin worst week five month crypto analyst s...</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>bitcoin worst week five month crypto analyst s...</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>bitcoin worst week five month crypto analyst s...</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>bitcoin worst week five month crypto analyst s...</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>bitcoin worst week five month crypto analyst s...</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77633</th>\n",
       "      <td>ted best dev space xgems crypto bitcoin</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77634</th>\n",
       "      <td>rolled taco feedgoat found bitcoin user vault ...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77635</th>\n",
       "      <td>psychojellybean found bitcoin user vault locat...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77636</th>\n",
       "      <td>buying bitcoin today could best decision ever</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77637</th>\n",
       "      <td>ponchovaldez found bitcoin user vault location...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>77638 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    Text  sentiment\n",
       "0      bitcoin worst week five month crypto analyst s...       -1.0\n",
       "1      bitcoin worst week five month crypto analyst s...       -1.0\n",
       "2      bitcoin worst week five month crypto analyst s...       -1.0\n",
       "3      bitcoin worst week five month crypto analyst s...       -1.0\n",
       "4      bitcoin worst week five month crypto analyst s...       -1.0\n",
       "...                                                  ...        ...\n",
       "77633            ted best dev space xgems crypto bitcoin        1.0\n",
       "77634  rolled taco feedgoat found bitcoin user vault ...        1.0\n",
       "77635  psychojellybean found bitcoin user vault locat...        1.0\n",
       "77636      buying bitcoin today could best decision ever        1.0\n",
       "77637  ponchovaldez found bitcoin user vault location...        1.0\n",
       "\n",
       "[77638 rows x 2 columns]"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lemmatizer = WordNetLemmatizer()\n",
    "stemmer = PorterStemmer() \n",
    "\n",
    "#adding a counter to check the progress of the algo while it runs\n",
    "global counter\n",
    "counter = 0\n",
    "def preprocess(sentence, stemming=False, lemmatizing=False):\n",
    "  global counter\n",
    "  counter += 1\n",
    "  if counter % 100 == 0:\n",
    "    pass\n",
    "    #print(counter)\n",
    "\n",
    "  #clean as much as possible, but not apply strong editing to the text, yet\n",
    "  sentence=str(sentence)\n",
    "  tokenizer = RegexpTokenizer(r'\\w+')\n",
    "\n",
    "  sentence = sentence.lower()\n",
    "  sentence=sentence.replace('{html}',\"\") \n",
    "  cleanr = re.compile('<.*?>')\n",
    "  cleantext = re.sub(cleanr, '', sentence)\n",
    "  rem_url=re.sub(r'http\\S+', '',cleantext)\n",
    "  rem_num = re.sub('[0-9]+', '', rem_url)\n",
    "  tokens = tokenizer.tokenize(rem_num)\n",
    "  \n",
    "  filtered_words = [w for w in tokens if len(w) > 2 if not w in stopwords.words('english')]\n",
    "  \n",
    "  if stemming == True and lemmatizing == False:\n",
    "    stem_words=[stemmer.stem(w) for w in filtered_words]\n",
    "    return \" \".join(stem_words)\n",
    "\n",
    "  if stemming == False and lemmatizing == True:\n",
    "    lemma_words=[lemmatizer.lemmatize(w) for w in filtered_words]\n",
    "    return \" \".join(lemma_words)\n",
    "\n",
    "  if stemming == True and lemmatizing == True:\n",
    "    stem_words=[stemmer.stem(w) for w in filtered_words]\n",
    "    lemma_words=[lemmatizer.lemmatize(w) for w in stem_words]\n",
    "    return \" \".join(lemma_words)\n",
    "  \n",
    "  #at the end of the algo we return filtered words\n",
    "  return \" \".join(filtered_words)\n",
    "\n",
    "#preprocess the sentiment text\n",
    "#df_sentiment['Text'] = df_sentiment['Text'].apply(lambda x: preprocess(x, stemming=True, lemmatizing=True))\n",
    "df_sentiment['Text'] = df_sentiment['Text'].apply(lambda x: preprocess(x, stemming=False, lemmatizing=True))\n",
    "df_sentiment"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9a66e2bc",
   "metadata": {},
   "source": [
    "Plot scores:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "id": "dc212e12",
   "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": [
    "plt.plot(df_sentiment[['sentiment']])\n",
    "plt.xlabel('Tweet ID')\n",
    "plt.ylabel('Sentiment')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5b73648d",
   "metadata": {},
   "source": [
    "Plot scores on donut chart:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "id": "23166825",
   "metadata": {},
   "outputs": [],
   "source": [
    "n_very_neg = len(df_sentiment[df_sentiment['sentiment'] < -0.25])\n",
    "n_neg = len(df_sentiment[(df_sentiment['sentiment'] < 0) & (df_sentiment['sentiment'] > -0.25)])\n",
    "n_neut = len(df_sentiment[df_sentiment['sentiment'] == 0])\n",
    "n_pos = len(df_sentiment[(df_sentiment['sentiment'] > 0) & (df_sentiment['sentiment'] < 0.25)])\n",
    "n_very_pos = len(df_sentiment[df_sentiment['sentiment'] > 0.25])\n",
    "n_groups = [n_very_neg, n_neg,n_neut,n_pos,n_very_pos]\n",
    "names = ['very negative', 'negative', 'neutral', 'positive','very positive']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "id": "92301a84",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.pie(n_groups, labels=names, colors=['darkred','red','yellow','limegreen','green'],\n",
    "        wedgeprops = { 'linewidth' : 7, 'edgecolor' : 'white' })\n",
    "my_circle=plt.Circle( (0,0), 0.7, color='white')\n",
    "p=plt.gcf()\n",
    "p.gca().add_artist(my_circle)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a1a24595",
   "metadata": {},
   "source": [
    "Get some descriptives:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "id": "7bd58bca",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mean score:\n",
      "0.07209058882973744\n",
      "std:\n",
      "0.22843928825378096\n",
      "N negative:\n",
      "12322\n",
      "N positive:\n",
      "28916\n"
     ]
    }
   ],
   "source": [
    "df_neg = df_sentiment[df_sentiment['sentiment'] < 0]\n",
    "df_pos = df_sentiment[df_sentiment['sentiment'] > 0]\n",
    "\n",
    "print('mean score:')\n",
    "print(df_sentiment['sentiment'].mean())\n",
    "\n",
    "print('std:')\n",
    "print(df_sentiment['sentiment'].std())\n",
    "\n",
    "print('N negative:')\n",
    "print(len(df_neg))\n",
    "\n",
    "print('N positive:')\n",
    "print(len(df_pos))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "44bedb0c",
   "metadata": {},
   "source": [
    "Extract the most common words found in both positive and negative positive reviews:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "id": "a9bb11bd",
   "metadata": {},
   "outputs": [],
   "source": [
    "positive_words = pd.DataFrame(\n",
    "    [dict(Counter(' '.join(df_pos['Text'].values.tolist()).split(' ')))]).T.sort_values(0, ascending=False)[0:100].index\n",
    "negative_words = pd.DataFrame(\n",
    "    [dict(Counter(' '.join(df_neg['Text'].values.tolist()).split(' ')))]).T.sort_values(0, ascending=False)[0:100].index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "id": "1a58fb2f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['bitcoin', 'crypto', 'btc', 'price', 'ftx', 'market', 'time', 'like',\n",
       "       'eth', 'exchange', 'cryptocurrency', 'people', 'money', 'new', 'get',\n",
       "       'news', 'one', 'buy', 'good', 'ethereum', 'blockchain', 'amp', 'year',\n",
       "       'block', 'right', 'would', 'see', 'best', 'need', 'coin', 'make',\n",
       "       'free', 'high', 'think', 'know', 'binance', 'first', 'still', 'many',\n",
       "       'much', 'last', 'could', 'real', 'say', 'value', 'top', 'low', 'way',\n",
       "       'latest', 'going', 'sat', 'want', 'world', 'day', 'even', 'nft',\n",
       "       'thing', 'trading', 'wallet', 'better', 'use', 'great', 'nfts', 'worth',\n",
       "       'gold', 'back', 'take', 'week', 'asset', 'xrp', 'network', 'never',\n",
       "       'follow', 'also', 'cash', 'bnb', 'change', 'long', 'hold', 'tweet',\n",
       "       'next', 'really', 'altcoin', 'lightning', 'hour', 'today', 'self',\n",
       "       'currency', 'mining', 'trust', 'current', 'point', 'bank', 'let',\n",
       "       'live', 'work', 'fiat', 'bottom', 'token', 'custody'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 113,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "positive_words"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "id": "dfe9a243",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['bitcoin', 'crypto', 'btc', 'price', 'ftx', 'eth', 'last', 'rsi',\n",
       "       'ethereum', 'exchange', 'market', 'like', 'people', 'long', 'money',\n",
       "       'time', 'average', 'one', 'cryptocurrency', 'year', 'lower', 'get',\n",
       "       'moving', 'buy', 'indicator', 'need', 'upper', 'amp', 'bollinger',\n",
       "       'week', 'bad', 'binance', 'say', 'make', 'think', 'cold', 'would',\n",
       "       'coin', 'wallet', 'going', 'usdt', 'know', 'news', 'still', 'see',\n",
       "       'day', 'take', 'even', 'blockchain', 'mean', 'fake', 'value', 'next',\n",
       "       'way', 'thing', 'weekly', 'daily', 'could', 'scam', 'month', 'storage',\n",
       "       'live', 'game', 'hard', 'asset', 'trust', 'shit', 'trading', 'every',\n",
       "       'gold', 'wrong', 'back', 'worst', 'never', 'bank', 'mining', 'world',\n",
       "       'want', 'saying', 'use', 'term', 'industry', 'change', 'also',\n",
       "       'million', 'fund', 'currency', 'fiat', 'low', 'block', 'usd', 'nothing',\n",
       "       'past', 'everyone', 'token', 'since', 'sat', 'new', 'hour', 'self'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "negative_words"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ee302188",
   "metadata": {},
   "source": [
    "### 5. Conclusion\n",
    "\n",
    "According to this analysis, with a mean sentiment of 0.07 (SD = 0.22) there is a slight but prevalent positive opinion in BTC tweets."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "id": "e838c911",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Other function to clean the tweets\n",
    "#def cleanTwt(twt):\n",
    "#    twt = re.sub(\"#bitcoin\", 'bitcoin', twt) # removes the '#' from bitcoin\n",
    "#    twt = re.sub(\"#Bitcoin\", 'Bitcoin', twt) # removes the '#' from Bitcoin\n",
    "#    twt = re.sub('#[A-Za-z0-9]+', '', twt) # removes any string with a '#'\n",
    "#    twt = re.sub('\\\\n', '', twt) # removes the '\\n' string\n",
    "#    twt = re.sub('https:\\/\\/\\S+', '', twt) # removes any hyperlinks\n",
    "#    return twt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "id": "5b86c502",
   "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>Text</th>\n",
       "      <th>cleaned_tweets</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>99991</th>\n",
       "      <td>Hver watt som brukes på mining, er energi som ...</td>\n",
       "      <td>Hver watt som brukes på mining, er energi som ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99992</th>\n",
       "      <td>If you are/were and will be surprised about th...</td>\n",
       "      <td>If you are/were and will be surprised about th...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99993</th>\n",
       "      <td>🦄 Don’t Dream It’s Over 🦄\\n✨ #Bitcoin #Ethereu...</td>\n",
       "      <td>🦄 Don’t Dream It’s Over 🦄✨ Bitcoin   ✨</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99994</th>\n",
       "      <td>Play brand new multi-coin game + free bitcoin ...</td>\n",
       "      <td>Play brand new multi-coin game + free bitcoin ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99995</th>\n",
       "      <td>La casa de #crypto\\nMoney Heist!\\n#FTX #Binanc...</td>\n",
       "      <td>La casa de Money Heist!     Bitcoin</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99996</th>\n",
       "      <td>yes exactly because it broke down its higher h...</td>\n",
       "      <td>yes exactly because it broke down its higher h...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99997</th>\n",
       "      <td>don’t understand what bitcoiners are celebrati...</td>\n",
       "      <td>don’t understand what bitcoiners are celebrati...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99998</th>\n",
       "      <td>Das Blockchain Trilemma! Dezentralität, Skalie...</td>\n",
       "      <td>Das Blockchain Trilemma! Dezentralität, Skalie...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99999</th>\n",
       "      <td>Bitcoin prostě není možné vypnout. Smiřte se s...</td>\n",
       "      <td>Bitcoin prostě není možné vypnout. Smiřte se s...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100000</th>\n",
       "      <td>Bitcoin is only the first. That’s all , it’s n...</td>\n",
       "      <td>Bitcoin is only the first. That’s all , it’s n...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                     Text  \\\n",
       "99991   Hver watt som brukes på mining, er energi som ...   \n",
       "99992   If you are/were and will be surprised about th...   \n",
       "99993   🦄 Don’t Dream It’s Over 🦄\\n✨ #Bitcoin #Ethereu...   \n",
       "99994   Play brand new multi-coin game + free bitcoin ...   \n",
       "99995   La casa de #crypto\\nMoney Heist!\\n#FTX #Binanc...   \n",
       "99996   yes exactly because it broke down its higher h...   \n",
       "99997   don’t understand what bitcoiners are celebrati...   \n",
       "99998   Das Blockchain Trilemma! Dezentralität, Skalie...   \n",
       "99999   Bitcoin prostě není možné vypnout. Smiřte se s...   \n",
       "100000  Bitcoin is only the first. That’s all , it’s n...   \n",
       "\n",
       "                                           cleaned_tweets  \n",
       "99991   Hver watt som brukes på mining, er energi som ...  \n",
       "99992   If you are/were and will be surprised about th...  \n",
       "99993              🦄 Don’t Dream It’s Over 🦄✨ Bitcoin   ✨  \n",
       "99994   Play brand new multi-coin game + free bitcoin ...  \n",
       "99995                 La casa de Money Heist!     Bitcoin  \n",
       "99996   yes exactly because it broke down its higher h...  \n",
       "99997   don’t understand what bitcoiners are celebrati...  \n",
       "99998   Das Blockchain Trilemma! Dezentralität, Skalie...  \n",
       "99999   Bitcoin prostě není možné vypnout. Smiřte se s...  \n",
       "100000  Bitcoin is only the first. That’s all , it’s n...  "
      ]
     },
     "execution_count": 145,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#df[\"Text\"] = df[\"Text\"].apply(lambda x : ' '.join([s for s in x.split(' ') if s.find('@') == -1 and s.find('www') == -1 and s.find('https') == -1]))\n",
    "#df['cleaned_tweets'] = df['Text'].apply(cleanTwt)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d1bfa2c4",
   "metadata": {},
   "outputs": [],
   "source": [
    "#detect_langs(DF2.loc[3, \"cleaned_tweets\"])"
   ]
  }
 ],
 "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}