Quantpylib GeneticRegression for Quant Trading Analysis
A few days ago, we announced the quantpylib Python library’s feature introduction to incorporate backtesting for crypto assets. We demonstrated how you can test ideas such as trend following under vol targeting frameworks, using just a few lines of code.
We are going to be introducing EVEN more exciting features into quantpylib over the next few days. Part of the PR has been made, and sits under quantpylib.simulator.models.
The feature is a regression tool, an abstraction written on top of Gene’s mathematical evaluator-parser and the popular statsmodels library to put in your hands a lightning-fast analytical tool for building regression models relating to market data.
We have already made this available to you in the backtesting-side, but backtesting is just one of the many tools in the quantitative research process. It is only natural that we exploit this power to for other steps in the research process such as regression analysis.
All you would have to do, to say, analyse a regression of forward-one day log returns against time-series momentum and cross-section momentum would be to specify…
forward_1(logret_1()) ~ tszscre_25(close) + csrank(logret_25())
And then voila the library will take care of all of the data manipulation and regression using statsmodels under the hood!
Sounds exciting?