Compute data-adaptive parameter estimate for a single cross-validation fold

cv_param_est(fold, data, parameter_wrapper, absolute, negative, n_top,
  learning_library, Y_name, A_name, W_name)

Arguments

fold

fold output from origami

data

entire training data

parameter_wrapper

user-defined function

absolute

boolean: TRUE = test for absolute effect size. This FALSE = test for directional effect. This overrides argument negative.

negative

boolean: TRUE = test for negative effect size, FALSE = test for positive effect size

n_top

integer value for the number of candidate covariates to generate using the data-adaptive estimation algorithm

learning_library

character of SuperLearner library

Y_name

(character) colnames of all biomarkers

A_name

(character) colnames of treatment

W_name

(character) colnames of all baseline covariates

Value

data_adaptive_index (integer vector) rank for each gene

index_grid (integer matrix) gene index from rank 1 to rank K

psi_est estimand of DE for rank 1 to rank K genes

EIC_est estimand of EIC for rank 1 to rank K genes