app.py

import tweepy
import time
import pandas as pd
from transformers import pipeline
import matplotlib.pyplot as plt
import gradio as gr
import os

def twitter_auth(consumerkey,consumersecret):
  consumer_key = consumerkey
  consumer_secret = consumersecret

  auth = tweepy.AppAuthHandler(consumer_key,consumer_secret)

  api = tweepy.API(auth,wait_on_rate_limit= True)
  return api

"""## Helper function for handling ratelimit and pagination"""

def limit_handled(cursor):
  """
  Function takes the cursor and returns tweets
  """
  while True:
    try:
      yield cursor.next()
    except tweepy.errors.TweepyException:
      print('reached rate limit, sleeping for > 15 mins')
      time.sleep(15*61)
    except StopIteration:
      break

def tweets_collector(query,count):
  consumerkey = os.environ.get('consumerkey')
  consumersecret = os.environ.get('consumersecret')
  api = twitter_auth(consumerkey,consumersecret)
  query = query +' -filter:retweets'
  search = limit_handled(tweepy.Cursor(api.search_tweets,q = query,tweet_mode = 'extended',lang ='en',result_type ='recent').items(count))
  sentiment_analysis = pipeline(model = "finiteautomata/bertweet-base-sentiment-analysis")
  tweets = []

  for tweet in search:
    try:
      content = tweet.full_text
      sentiment = sentiment_analysis(content)
      tweets.append({'tweet' : content ,'sentiment': sentiment[0]['label']})
    except:
      pass
  return tweets

"""## Run sentiment Analysis"""

#tweets = tweets_collector(query,count)
#df = pd.DataFrame(tweets)

import pandas as pd

pd.set_option('max_colwidth',None)
pd.set_option('display.width',3000)

#import matplotlib.pyplot as plt

#sentiment_counts = df.groupby(['sentiment']).size()

#fig = plt.figure(figsize = (6,6),dpi = 100)
#ax = plt.subplot(111)
#sentiment_counts.plot.pie(ax = ax,autopct = '%1.f%%',startangle = 270,fontsize = 12,label = "")

def complaint_analysis(query,count):
  tweets = tweets_collector(query,count)
  df = pd.DataFrame(tweets)
  from wordcloud import WordCloud
  from wordcloud import STOPWORDS
  sentiment_counts = df.groupby(['sentiment']).size()
  fig = plt.figure(figsize = (6,6),dpi = 100)
  ax = plt.subplot(111)
  sentiment_counts.plot.pie(ax = ax,autopct = '%1.f%%',startangle = 270,fontsize = 12,label = "")
  plt.savefig('Overall_satisfaction.png')

  positive_tweets = df['tweet'][df['sentiment'] == 'POS']
  stop_words = ["https","co","RT","ola_supports","ola_cabs","customer"] + list(STOPWORDS)
  positive_wordcloud = WordCloud(max_font_size=50,max_words = 30,background_color="white",stopwords=stop_words).generate(str(positive_tweets))
  plt.figure()
  plt.title("Positive Tweets - Wordcloud")
  plt.imshow(positive_wordcloud,interpolation="bilinear")
  plt.axis("off")
  #plt.show()
  plt.savefig('positive_tweet.png')
  negative_tweets = df['tweet'][df['sentiment'] == 'NEG']
  stop_words = ["https","co","RT","ola_supports","ola_cabs","customer"] + list(STOPWORDS)
  negative_wordcloud = WordCloud(max_font_size=50,max_words = 30,background_color="white",stopwords=stop_words).generate(str(negative_tweets))
  plt.figure()
  plt.title("Negative Tweets - Wordcloud")
  plt.imshow(negative_wordcloud,interpolation="bilinear")
  plt.axis("off")
  #plt.show()
  plt.savefig('negative_tweet.png')
  return ['Overall_satisfaction.png','positive_tweet.png','negative_tweet.png']

gr.Interface(fn=complaint_analysis, 
             inputs=[
            gr.inputs.Textbox(
                placeholder="Tweet handle please", label="Company support Twitter Handle", lines=5), gr.Slider(100, 1000) ],
             outputs= [gr.outputs.Image(type="pil"),gr.outputs.Image(type="pil"),gr.outputs.Image(type="pil")],
             examples=[]).launch(debug= True)