app.py

import streamlit as st
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import nltk
from textblob import TextBlob
import numpy as np
from wordcloud import WordCloud
import plotly.express as px
from datetime import datetime, timedelta
from transformers import pipeline
from vaderSentiment.vaderSentiment import SentimentIntensityAnalyzer
from sklearn.linear_model import LinearRegression
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import LabelEncoder

# Ensure necessary NLTK data is available
nltk.download('punkt')

st.set_page_config(page_title="Advanced Sentiment Analyzer", layout="wide")

st.title("🔍 Advanced Sentiment Analysis Dashboard")
st.markdown("Analyze sentiments with deep insights and visualizations")

# Sidebar for user input
st.sidebar.header("Upload your dataset")
uploaded_file = st.sidebar.file_uploader("Upload a CSV file", type=["csv"])

if uploaded_file is not None:
    df = pd.read_csv(uploaded_file)
    st.sidebar.success("File uploaded successfully!")
else:
    st.sidebar.warning("Please upload a CSV file to proceed.")

# Sample Text Input
st.sidebar.subheader("Enter Text for Sentiment Analysis")
user_input = st.sidebar.text_area("Type or paste text here", "The product is amazing!")

# Initialize sentiment analyzers
analyzer = SentimentIntensityAnalyzer()
bert_sentiment = pipeline("sentiment-analysis")

def analyze_vader_sentiment(text):
    score = analyzer.polarity_scores(text)['compound']
    return "Positive" if score > 0.05 else "Negative" if score < -0.05 else "Neutral"

def analyze_bert_sentiment(text):
    result = bert_sentiment(text)[0]
    return result['label']

if user_input:
    vader_result = analyze_vader_sentiment(user_input)
    bert_result = analyze_bert_sentiment(user_input)
    st.sidebar.markdown(f"**VADER Sentiment:** {vader_result}")
    st.sidebar.markdown(f"**BERT Sentiment:** {bert_result}")

# If dataset is uploaded, analyze sentiments
if uploaded_file is not None:
    df['VADER Sentiment'] = df['text'].apply(analyze_vader_sentiment)
    df['BERT Sentiment'] = df['text'].apply(analyze_bert_sentiment)
    
    # Sentiment Distribution
    st.subheader("📊 Sentiment Distribution")
    fig, ax = plt.subplots()
    sns.countplot(x=df['VADER Sentiment'], palette="coolwarm", ax=ax)
    st.pyplot(fig)
    
    # Word Cloud
    st.subheader("☁️ Word Cloud")
    sentiment_filter = st.selectbox("Filter Word Cloud by Sentiment", ["All", "Positive", "Neutral", "Negative"])
    if sentiment_filter != "All":
        filtered_df = df[df['VADER Sentiment'] == sentiment_filter]
    else:
        filtered_df = df
    all_text = " ".join(filtered_df['text'].dropna().astype(str))
    wordcloud = WordCloud(width=800, height=400, background_color='white').generate(all_text)
    st.image(wordcloud.to_array(), use_column_width=True)
    
    # Sentiment Over Time (Simulated for Demo)
    st.subheader("📈 Sentiment Trend Over Time")
    df['date'] = pd.date_range(end=datetime.today(), periods=len(df), freq='D')
    sentiment_map = {'Positive': 1, 'Neutral': 0, 'Negative': -1}
    df['sentiment_score'] = df['VADER Sentiment'].map(sentiment_map)
    
    fig = px.line(df, x='date', y='sentiment_score', title='Sentiment Trend')
    st.plotly_chart(fig)
    
    # Regression Model for Sentiment Prediction
    st.subheader("📈 Sentiment Prediction")
    label_encoder = LabelEncoder()
    df['sentiment_encoded'] = label_encoder.fit_transform(df['VADER Sentiment'])
    X = np.array(range(len(df))).reshape(-1, 1)
    y = df['sentiment_encoded']
    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
    model = LinearRegression()
    model.fit(X_train, y_train)
    future_days = np.array(range(len(df), len(df) + 30)).reshape(-1, 1)
    predicted_sentiments = model.predict(future_days)
    
    fig_pred = px.line(x=range(len(df), len(df) + 30), y=predicted_sentiments, title='Predicted Sentiment Trend')
    st.plotly_chart(fig_pred)
    
    # Interactive Pie Chart
    st.subheader("📝 Sentiment Breakdown")
    sentiment_counts = df['VADER Sentiment'].value_counts()
    fig_pie = px.pie(names=sentiment_counts.index, values=sentiment_counts.values, title="Sentiment Breakdown")
    st.plotly_chart(fig_pie)
    
    # Sentiment Heatmap
    st.subheader("🔥 Sentiment Heatmap")
    heatmap_df = df.pivot_table(index=df['date'].dt.date, columns='VADER Sentiment', aggfunc='size', fill_value=0)
    fig_heatmap = px.imshow(heatmap_df.T, title="Sentiment Heatmap", labels=dict(x="Date", y="Sentiment", color="Count"))
    st.plotly_chart(fig_heatmap)
    
    # Keyword Sentiment Analysis
    st.subheader("🔍 Keyword Sentiment Analysis")
    keyword = st.text_input("Enter a keyword to analyze sentiment", "great")
    keyword_df = df[df['text'].str.contains(keyword, case=False, na=False)]
    st.write(keyword_df[['text', 'VADER Sentiment']])
    
    # Download Report as CSV
    st.subheader("📄 Download Report")
    csv = df.to_csv(index=False).encode('utf-8')
    st.download_button(label="Download CSV", data=csv, file_name="sentiment_analysis.csv", mime='text/csv')

st.sidebar.markdown("Developed with ❤️")