Projection Predictive Feature Selection


[Up] [Top]

Documentation for package ‘projpred’ version 2.2.1

Help Pages

projpred-package Projection predictive feature selection
as.matrix.projection Extract projected parameter draws
break_up_matrix_term Break up matrix terms
cl_agg Weighted averaging within clusters of parameter draws
cv-indices Create cross-validation folds
cvfolds Create cross-validation folds
cv_ids Create cross-validation folds
cv_varsel Variable selection with cross-validation
cv_varsel.default Variable selection with cross-validation
cv_varsel.refmodel Variable selection with cross-validation
df_binom Binomial toy example
df_gaussian Gaussian toy example
extend_family Extend a family
extra-families Extra family objects
get_refmodel Reference model structure
get_refmodel.default Reference model structure
get_refmodel.refmodel Reference model structure
get_refmodel.stanreg Reference model structure
get_refmodel.vsel Reference model structure
init_refmodel Reference model structure
mesquite Mesquite data set
plot.vsel Plot summary statistics of a variable selection
pred-projection Predictions from a submodel (after projection)
predict.refmodel Predictions or log predictive densities from a reference model
print.vsel Print results (summary) of variable selection
print.vselsummary Print summary of variable selection
project Projection onto submodel(s)
projpred Projection predictive feature selection
proj_linpred Predictions from a submodel (after projection)
proj_predict Predictions from a submodel (after projection)
refmodel-init-get Reference model structure
solution_terms Retrieve predictor solution path or predictor combination
solution_terms.projection Retrieve predictor solution path or predictor combination
solution_terms.vsel Retrieve predictor solution path or predictor combination
Student_t Extra family objects
suggest_size Suggest submodel size
suggest_size.vsel Suggest submodel size
summary.vsel Summary statistics of a variable selection
varsel Variable selection (without cross-validation)
varsel.default Variable selection (without cross-validation)
varsel.refmodel Variable selection (without cross-validation)