Find past market days similar to today and see what actually happened next.
pastdays is a Python library that compares current partial market data with historical data to identify similar past trading days and analyze their outcomes.
It does not predict, does not trade, and does not generate signals. It simply answers one question:
“When the market looked like this before, what happened afterwards?”
Most backtests fail because they assume: - perfect execution - knowledge of the future - curve-fitted parameters
Traders, however, think differently:
“Have I seen this kind of day before?”
pastdays is built around that mindset — focusing on historical context, not forecasts.
- Accepts **user-provided historical data**
- Accepts **current-day partial data**
- Extracts **simple, explainable features**
- Finds **K most similar historical days**
- Reports **what happened next** on those days
No machine learning.
No black box.
pip install pastdays