📅 2024-01-18 — Session: Enhanced OLS Residuals Calculation and Visualization
🕒 16:05–17:10
🏷️ Labels: Python, OLS, Data Visualization, Rolling Residuals, Finance
📂 Project: Dev
⭐ Priority: MEDIUM
Session Goal
The session aimed to enhance the calculation and visualization of Ordinary Least Squares (OLS) residuals for financial data analysis.
Key Activities
- Modified Function for Rolling Residuals Calculation: Updated the
add_rolling_residuals
function to perform OLS computations directly within the function, allowing for better debugging and control. - Line Plot Creation: Developed a line plot of adjusted close prices with residuals, due to the unavailability of
mplfinance
for candlestick charts. - Diagnostic Plot for OLS Model: Outlined the steps to create a diagnostic plot for an OLS model, including regression line, residuals, and model parameter annotations.
- Integration with Candlestick Chart: Provided a script to integrate OLS diagnostic plots with candlestick charts, enhancing stock data visualization.
- Efficient Calculation for Multiple Tickers: Implemented a script to apply the rolling residuals function across multiple tickers and save the results to a CSV file.
Achievements
- Successfully modified and tested the
add_rolling_residuals
function. - Created visualizations that integrate OLS diagnostics with stock data, improving analytical insights.
Pending Tasks
- Explore alternatives or updates for
mplfinance
to support candlestick charting directly. - Further optimize the rolling residuals calculation for large datasets.