Sparse Additive Modelling

Computationally efficient tools for high dimensional predictive modeling (regression and classification). SAM is short for sparse additive modeling, and adopts the computationally efficient basis spline technique. We solve the optimization problems by various computational algorithms including the block coordinate descent algorithm, fast iterative soft-thresholding algorithm, and newton method. The computation is further accelerated by warm-start and active-set tricks.

SAM

Sparse Additive Modelling (SAM) with:

• R package implementation (R/, src/, man/)
• Python wrapper package (python-package/) that reuses the current C++ core

R package

Build/check:

R CMD build .
R CMD check --as-cran SAM_1.2.tar.gz

Python package (wrapper)

Location: python-package/

APIs implemented:

• samLL, samHL, samEL, samQL
• predict_samLL, predict_samHL, predict_samEL, predict_samQL
• path plotting and summary helpers

The wrapper calls native symbols from src/SAM.so:

• grpLR, grpSVM, grpPR, grplasso

Quick local run:

bash python-package/scripts/build_native.sh
PYTHONPATH=python-package python3 python-package/examples/run_smoke.py

Documentation

• R help files: man/*.Rd
• Python docs: python-package/docs/ + python-package/mkdocs.yml

To build Python docs locally:

cd python-package
mkdocs build --strict