How virtrivedi/stock-price-modeling-analytics works
Project type and purpose
• From the file tree and README title “Stock-Price-Modeling-Analytics,” this is a data-analytics pipeline focused on processing market data, computing correlations, generating price features, and producing visualizations and model outputs. It is not a client-server or web application; rather it’s a collection of standalone scripts that form a processing pipeline for stock-price analysis.
File structure
• Top-level Python scripts:
• bar_generation.py – Aggregates tick data into time bars (e.g., minute/second OHLC).
• correlation_generation.py – Computes correlation matrices from price or feature series.
• correlation_heatmap.py – Reads correlation results and plots a heatmap.
• parse_book_fills.py – Parses trade‐fill events from raw order‐book data.
• parse_book_tops.py(.cpp/.py) – Parses top‐of-book snapshots; C++ for performance, Python as fallback.
• price_correlation.py – A wrapper or specialized correlation calculator.
• price_prediction.py – Trains or applies predictive models on engineered features.
• No dedicated folders, but implicit modular separation by functionality: data ingestion/parsing, feature engineering, modeling, visualization.
README for additional insights
• The README is sparse, so infer architecture from filenames: a linear pipeline that starts with raw data parsing, proceeds through feature generation, then correlation analysis and visualization, finishing with price‐prediction modeling.
Detailed capturing of architectural elements
• Parallel downstream stages: one branch feeding correlation_heatmap, another feeding price_prediction.
[Data Sources] → [parse_book_tops (C++)] → [parsed order-book data]
↘
→ [parse_book_fills (Py)] → [parsed fills]
→ [bar_generation] → [OHLC bar CSV]
→ [correlation_generation / price_correlation] → [correlation matrix]
↘
→ [correlation_heatmap]
→ [price_prediction] → [model outputs / performance metrics]