gbencke/pyAstroTrader ? reverse-engineered prompt

Build me a small research app that explores whether astrology lines up with stock price trends.

I want it to let me pick a ticker symbol, pull about 20 years of daily price data from AlphaVantage, calculate the planetary positions for each day, then train a machine learning model to look for patterns between astrological events and whether the price goes up or down over the next few days. After that, I want a separate step that loads the trained model and generates predictions for the next 180 days, saving the results so I can review them later.

Please make it easy to run from notebooks, with clear steps for downloading data, building the model, and making predictions. It should save input and output files in simple folders, and I should be able to start everything locally with one script. If you need current docs for any library or API details, look them up online.