import pandas as pd

import numpy as np

from textblob import TextBlob

import emoji

import matplotlib.pyplot as plt

import seaborn as sns

import warnings; warnings.simplefilter('ignore')

from transformers import pipeline

from langdetect import detect

from googletrans import Translator

from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score,confusion_matrix

df_emoji = pd.read_csv("Emoji_Sentiment_Data.csv",usecols = ['Emoji', 'Negative', 'Neutral', 'Positive'])

df_emoji.head(3)

polarity_list = []

for index, row in df_emoji.iterrows():

polarity = 0

arg_1 = row['Positive'] > row['Negative']

arg_2 = row['Positive'] == row['Negative'] and row['Neutral'] % 2 != 0

if arg_1 or arg_2:

polarity = 1

polarity_list.append(polarity)

new_df_emoji = pd.DataFrame(polarity_list, columns=['sentiment'])

new_df_emoji['emoji'] = df_emoji['Emoji'].values

new_df_emoji.head(3)

# print out the emoticons and sentiment values

e_c, p = 0, 0

for index, row in new_df_emoji.iterrows():

print(f"{row['emoji']} = {row['sentiment']}")

p += 1 if row['sentiment'] else 0

e_c += 1

print(f'Total Positive Emojis are ({p}:{e_c}) or {round(p / e_c * 100)}%')

text = "i ❤ sentiments #goodlife @today"

def extract_text_and_emoji(text = text):

global allchars, emoji_list

remove_keys = ('@', 'http://','https://', '&', '#')

clean_text = ' '.join(txt for txt in text.split() if not txt.startswith(remove_keys))

print(clean_text)

allchars = [str for str in text]

emoji_list = [c for c in allchars if c in emoji.EMOJI_DATA]

clean_text = ' '.join([str for str in clean_text.split() if not any(i in str for i in emoji_list)])

clean_emoji = ''.join([str for str in text.split() if any(i in str for i in emoji_list)])

return (clean_text, clean_emoji)

allchars, emoji_list = 0, 0

(ct, ce) = extract_text_and_emoji()

print('\nAll Char:', allchars)

print('\nAll Emoji:',emoji_list)

print('\n', ct)

print('\n',ce)

def detect_and_translate2(text):

translator = Translator()

lang = translator.detect(text).lang

trans = translator.translate(text, src=lang, dest='en')

return trans.text

# Load pre-trained sentiment analysis pipeline

sentiment_pipeline = pipeline('sentiment-analysis')

def get_sentiment(s_input='i hate sentiment analysis'):

results = sentiment_pipeline(s_input)

pred_senti = results[0]['label']

if pred_senti == 'POSITIVE':

pc = 1

else:

pc = 0

return pc

print(get_sentiment())

def get_emoji_sentiment(emoji_ls = '❤❤', emoji_df = new_df_emoji):

emoji_val_ls = []

for e in emoji_ls:

get_emo_senti = [row['sentiment'] for index, row in emoji_df.iterrows() if row['emoji'] == e]

if get_emo_senti: # Check if the list is not empty

emoji_val_ls.append(get_emo_senti[0])

else:

emoji_val_ls.append(0) # Or any default value you prefer

return emoji_val_ls

ges = get_emoji_sentiment()

print('Sentiment value of each emoji:',ges)

def analyze_emotion(text):

blob = TextBlob(text)

sentiment = blob.sentiment

if sentiment.polarity > 0:

if sentiment.subjectivity > 0.5:

return "joy" # Happy

else:

return "surprise" # Surprise

elif sentiment.polarity < 0:

if sentiment.subjectivity > 0.5:

return "anger" # Anger

else:

return "sadness" # Sad

else:

return "neutral" # Neutral

tweet = "Had a terrible experience at the restaurant tonight. Never going back."

emotion = analyze_emotion(tweet)

print(f"Emotion: {emotion}")

def get_text_emoji_sentiment(input_test = 'i hate 😒 sentiment analysis'):

(ext_text, ext_emoji) = extract_text_and_emoji(input_test)

print(f'Extracted: "{ext_text}" , {ext_emoji}')

ttext=""

lang = detect(ext_text)

if lang!="en":

ext_text= detect_and_translate2(ext_text)

print(f'Translated: "{ext_text}"')

senti_text = get_sentiment(ext_text)

print(f'Text value: {senti_text}')

senti_emoji_value = sum(get_emoji_sentiment(ext_emoji, new_df_emoji))

print_emo_val_avg = 0 if len(ext_emoji) == 0 else senti_emoji_value/len(ext_emoji)

print(f'Emoji average value: {print_emo_val_avg}')

senti_avg = (senti_emoji_value + senti_text) / (len(ext_emoji) + 1)

print(f'Average value: {senti_avg}')

senti_truth = "Positive" if senti_avg >= 0.5 else "Negative"

emtext = analyze_emotion(ext_text)

print(f'Text Emotion: {emtext}')

return senti_truth

print(get_text_emoji_sentiment())

test_df = pd.read_csv("test.csv",usecols = ['text', 'label'])

test_df=test_df[test_df['label'] != 'neutral']

test_df['label'] = test_df['label'].apply(lambda x: 1 if x == 'positive' else 0)

test_df.head(6)

def preprocess_text(text):

text = text.lower()

text = ' '.join(word for word in text.split() if not word.startswith(('@', 'http', '#')))

return text if text else " "

test_df['text'] = test_df['text'].apply(preprocess_text)

def detect_and_translate3(text):

if not text.strip():

return " "

translator = Translator()

try:

lang = detect(text)

if lang != 'en':

trans = translator.translate(text, src=lang, dest='en')

return trans.text

except:

pass

return text

def get_sentiment(text):

text= detect_and_translate3(text)

result = sentiment_pipeline(text)[0]['label']

return 1 if result == 'POSITIVE' else 0

test_df['predicted_sentiment'] = test_df['text'].apply(get_sentiment)

y_true = test_df['label']

y_pred = test_df['predicted_sentiment']

accuracy = accuracy_score(y_true, y_pred)

precision = precision_score(y_true, y_pred)

recall = recall_score(y_true, y_pred)

f1 = f1_score(y_true, y_pred)

print("Accuracy ", accuracy)

print("Precision ", precision)

print("Recall ", recall)

print("F1 Score ", f1)

cm = confusion_matrix(y_true, y_pred)

plt.figure(figsize=(8, 6))

sns.heatmap(cm, annot=True, fmt='d', cmap='Blues', cbar=False,

xticklabels=['Negative', 'Positive'], yticklabels=['Negative', 'Positive'])

plt.xlabel('Predicted Label')

plt.ylabel('True Label')

plt.title('Confusion Matrix')

plt.show()