SPSP 0.2.0

New features

  1. Add the supports to the SCAD-penalized and MCP-penalized regression model through the fitfun.SP = SCAD.ncvreg, fitfun.SP = MCP.ncvreg. The previous established methods are ridge.glmnet. lasso.glmnet, adalasso.glmnet, and adalassoCV.glmnet.

SPSP 0.1.1

patch changes

  1. Fix the function corresponding to the interceptargument. The intercept argument controls whether to include the intercept, but since the SPSP algorithm estimates the intercept separately. It needs to be dealt with separately.

  2. The standardize only control the standardization of covariates.

SPSP 0.1.0

Major changes

  1. Implement the Selection by Partitioning the Solution Paths procedure in the SPSP() and SPSP_step() functions;

  2. Add fitfun.SP functions to obtain the solution paths for the SPSP algorithm;

  3. Include one high-dimensional data for illustration purpose.