task-3 module added

scripts/task.py

@@ -0,0 +1,72 @@
+import sys
+import os
+
+# Add src directory to the Python path
+src_path = os.path.abspath(os.path.join(os.path.dirname(__file__), '../src'))
+sys.path.append(src_path)
+
+from data_loader import load_data, load_and_concatenate_stock_data
+from data_cleaner import clean_data, drop_first_column, format_date_column
+from sentiment_analysis import analyze_sentiment, aggregate_sentiments_by_date
+from stock_returns import calculate_daily_returns
+from data_merger import merge_data, merge_sentiment_with_stock
+from correlation_analysis import calculate_correlation
+from visualizationt import plot_daily_returns, scatter_plot
+
+
+def main():
+    # Load data
+    news_df = load_data('../data/news.csv')
+
+    # File paths and corresponding company names
+    file_paths = [
+        '../data/apple.csv',
+        '../data/amazon.csv',
+        '../data/google.csv',
+        '../data/meta.csv',
+        '../data/microsoft.csv',
+        '../data/nvidia.csv',
+        '../data/tesla.csv'
+    ]
+
+    company_names = ['Apple', 'Amazon', 'Google', 'Meta', 'Microsoft', 'Nvidia', 'Tesla']
+
+    # Load and concatenate the stock data
+    stock_df = load_and_concatenate_stock_data(file_paths, company_names)
+
+    # Drop the first column
+    news_df = drop_first_column(news_df)
+
+    # Clean data
+    stock_df = clean_data(stock_df, 'Date')
+    news_df = clean_data(news_df, 'date')
+
+    # Format the 'date' columns
+    news_df = format_date_column(news_df, 'date')
+    stock_df = format_date_column(stock_df, 'date')
+
+    # Merge data
+    merged_df = merge_data(news_df, stock_df)
+
+    # Analyze sentiment
+    merged_df['sentiment'] = merged_df['headline'].apply(analyze_sentiment)
+
+    # Calculate daily stock returns
+    merged_df = calculate_daily_returns(merged_df)
+
+    # Aggregate sentiments by date
+    daily_sentiment = aggregate_sentiments_by_date(merged_df)
+
+    # Merge sentiment data with stock data
+    daily_df = merge_sentiment_with_stock(merged_df, daily_sentiment)
+
+    # Calculate correlation
+    correlation = calculate_correlation(daily_df)
+    print(f'Pearson correlation coefficient: {correlation}')
+
+    # Visualize
+    plot_daily_returns(daily_df)
+    scatter_plot(daily_df)
+
+if __name__ == '__main__':
+    main()

src/correlation_analysis.py

@@ -0,0 +1,6 @@
+
+import pandas as pd
+
+def calculate_correlation(df, column1='Average Sentiment', column2='Daily Return'):
+    correlation = df[[column1, column2]].corr().iloc[0, 1]
+    return correlation

src/data_cleaner.py

@@ -0,0 +1,24 @@
+import pandas as pd
+
+def clean_data(df, date_column):
+    # Check if the column name needs to be renamed to 'date'
+    if date_column in df.columns:
+        df.rename(columns={date_column: 'date'}, inplace=True)
+
+    # Convert the date column to datetime and remove timezone information
+    df['date'] = pd.to_datetime(df['date'], errors='coerce').dt.tz_localize(None)
+
+    # Drop rows with any missing values
+    df.dropna(inplace=True)
+
+    return df
+
+def drop_first_column(df):
+
+    df = df.drop(df.columns[0], axis=1)
+    return df
+
+def format_date_column(df, date_column):
+
+    df[date_column] = pd.to_datetime(df[date_column], errors='coerce').dt.strftime('%Y-%m-%d')
+    return df

src/data_loader.py

@@ -0,0 +1,23 @@
+import pandas as pd
+
+def load_data(file_path):
+    return pd.read_csv(file_path)
+
+def load_and_concatenate_stock_data(file_paths, company_names):
+
+    # List to store individual dataframes
+    dataframes = []
+
+    # Loop through each file path and company name
+    for file_path, company_name in zip(file_paths, company_names):
+        # Load the CSV file
+        df = pd.read_csv(file_path)
+        # Add a 'Company' column
+        df['Company'] = company_name
+        # Append the dataframe to the list
+        dataframes.append(df)
+
+    # Concatenate all dataframes into one
+    stock_df = pd.concat(dataframes)
+
+    return stock_df

src/data_merger.py

@@ -0,0 +1,9 @@
+import pandas as pd
+
+def merge_data(news_df, stock_df, on='date'):
+    merged_df = pd.merge(news_df, stock_df, on=on)
+    return merged_df
+
+def merge_sentiment_with_stock(stock_df, sentiment_df):
+    daily_df = pd.merge(stock_df[['date', 'Daily Return']], sentiment_df, on='date', how='inner')
+    return daily_df

src/sentiment_analysis.py

@@ -0,0 +1,12 @@
+
+import pandas as pd
+from textblob import TextBlob
+
+def analyze_sentiment(text):
+    analysis = TextBlob(text)
+    return analysis.sentiment.polarity
+
+def aggregate_sentiments_by_date(df, date_column='date', sentiment_column='sentiment', output_column='Average Sentiment'):
+    daily_sentiment = df.groupby(date_column)[sentiment_column].mean().reset_index()
+    daily_sentiment.columns = [date_column, output_column]
+    return daily_sentiment

src/stock_returns.py

@@ -0,0 +1,6 @@
+
+import pandas as pd
+
+def calculate_daily_returns(df, close_column='Close', return_column='Daily Return'):
+    df[return_column] = df[close_column].pct_change() * 100
+    return df

src/visualizationt.py

@@ -0,0 +1,32 @@
+
+import matplotlib.pyplot as plt
+import seaborn as sns
+
+def plot_daily_returns(daily_df):
+    plt.figure(figsize=(14, 7))
+
+    plt.subplot(2, 1, 1)
+    plt.plot(daily_df.index, daily_df['Daily Return'], label='Daily Return', color='blue')
+    plt.title('Daily Stock Returns')
+    plt.xlabel('Date')
+    plt.ylabel('Return (%)')
+    plt.legend()
+
+    plt.subplot(2, 1, 2)
+    plt.plot(daily_df.index, daily_df['Average Sentiment'], label='Average Sentiment', color='orange')
+    plt.title('Average Daily Sentiment')
+    plt.xlabel('Date')
+    plt.ylabel('Sentiment Score')
+    plt.legend()
+
+    plt.tight_layout()
+    plt.show()
+
+def scatter_plot(daily_df):
+    plt.figure(figsize=(14, 7))
+    sns.scatterplot(data=daily_df, x='Average Sentiment', y='Daily Return', alpha=0.5)
+    plt.title('Sentiment Score vs. Stock Returns')
+    plt.xlabel('Average Sentiment Score')
+    plt.ylabel('Daily Return (%)')
+    plt.legend()
+    plt.show()