Skip to content
Pre-release

@jason-d-jones jason-d-jones released this Dec 2, 2018 · 2 commits to master since this release

Installation

  • Select gauss-glmnet.zip with GAUSS (Tools > Install Application)
  • Use gpkg utility in GAUSS home directory (./gpkg install gauss-glmnet.zip).
  • See Installing GAUSS Application Modules for more details.

Supports

  • Gaussian distributed response variables (family = "normal").
  • Elasticnet mixing parameter.
  • Number of lambda values.
  • Observation weights.
  • Relative penalites for each predictor.
  • User specified lambda values.
  • Convergence threshold.
  • Maximum number of variables in the model.
  • Control over smallest lambda value and lambda spacing.
  • Variables (columns of X) to exclude.

Not yet supported

  • Response variables with binomial, poisson, multinomial and multivariate Gaussian.
  • Sparse matrix inputs.
  • Ability to specify that variables should NOT be standardized prior to fitting.

Known issues

  • None.
Assets 3
You can’t perform that action at this time.