Software
I’ve written some software packages that people can download. Here they are in list form.
- py-earth
- A Python implementation of the MARS algorithm written mostly in Cython, compatible with many prominent members of the Python data stack including pandas, scikit-learn, and patsy.
- choldate
- A Python package for performing rank 1 updates and downdates on Cholesky factors. Because why \(O\left( n^3 \right)\) when you can \(O\left( n^2 \right)\)? It works with numpy arrays from the C side, so it’s pretty fast.
- glm-sklearn
- A wrapper to make statsmodels’ GLM module compatible with scikit-learn. Useful if you want to put a GLM in a Pipeline, which, I admit, I have sometimes wanted to do.
- CLSOCP
- An SOCP solver written in R, using a smoothing Newton method technique.
- CONOR
- An R package for cross-platform normalization of microarray data, from back in my thesis days, you know.