GitHub - kgreenewald/teralasso

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 2 Commits
FFKron3.m		FFKron3.m
MCP_derivative.m		MCP_derivative.m
README		README
SCAD_derivative.m		SCAD_derivative.m
auc_comp.m		auc_comp.m
experiment1_lambdaSweep.m		experiment1_lambdaSweep.m
experiment2_numConvergence.m		experiment2_numConvergence.m
generate_cov.m		generate_cov.m
generate_cov_grid.m		generate_cov_grid.m
project_ksum.m		project_ksum.m
shrinkL1.m		shrinkL1.m
shrink_regularizer.m		shrink_regularizer.m
teralasso.m		teralasso.m
windExperiment.m		windExperiment.m

Repository files navigation

README

Main TeraLasso algorithm is teralasso.m

Three experiments are included as scripts: 

experiment1lambdaSweep.m: recreates Figure 8(b) with fewer random trials (to control runtime).

experiment2numConvergence.m: An experiment testing numerical convergence of the TeraLasso with nonconvex regularization, typical of those in Figure 15 (supplement).

windExperiment.m: Experiment recreating Figure 12. Our results are shown, the BiGLasso method is included but commented out due to its excessive runtime.