run_pipeline.py

"""
Pipeline Runner
===============
Runs the full earnings sentiment analysis pipeline end-to-end:

  Stage 1: Collect earnings transcripts (SEC EDGAR + synthetic fallback)
  Stage 2: Score transcripts with LM lexicon and VADER
  Stage 3: Run event study and statistical analysis

Run this file to execute the complete pipeline in one command.

Author: Atrija Haldar
"""

import sys
import os

sys.path.insert(0, os.path.dirname(__file__))

from collection.fetcher       import collect_transcripts, TICKERS
from sentiment.analyser       import score_transcripts, ticker_sentiment_summary
from analysis.price_reaction  import (fetch_price_data, run_event_study,
                                       analyse_sentiment_return_relationship,
                                       build_summary_table, build_dashboard,
                                       export_analysis_outputs, MARKET_PROXY)

def run():
    print("=" * 60)
    print("  Earnings Sentiment & Price Reaction Pipeline")
    print("=" * 60)

    print("\n[Stage 1/3] Collecting transcripts...")
    transcripts = collect_transcripts(TICKERS, quarters=8)

    print("\n[Stage 2/3] Scoring sentiment...")
    scores  = score_transcripts(transcripts)
    summary = ticker_sentiment_summary(scores)

    print("\n[Stage 3/3] Running event study & analysis...")
    stock_returns, market_returns = fetch_price_data(TICKERS, MARKET_PROXY)
    events        = run_event_study(scores, stock_returns, market_returns)
    stats_results = analyse_sentiment_return_relationship(events)
    stat_summary  = build_summary_table(events, stats_results)
    build_dashboard(events, stats_results, stat_summary)
    export_analysis_outputs(events, stats_results, stat_summary)

    print("\n" + "=" * 60)
    print("  Pipeline complete. Outputs saved to /output")
    print("  Open output/analysis/sentiment_analysis_dashboard.html")
    print("=" * 60)

if __name__ == "__main__":
    run()