tsmvr_gridsearch: Gridsearch for tsmvr

In spcorum/tsmvr-saved: Truly Sparse Multivariate Regression

Description

Calculates the mean and standard deviation of fold-averaged errors derived from replicated tsmvr k-fold cross-validation experiments. The error for each base model is the normalized squared error between the true response and the predicted response on a given cross-validation set.

Usage

tsmvr_gridsearch(X, Y, s1_vec, s2_vec, pars, quiet = F, seed = NULL)

Arguments

X	design matrix (n-by-p)
Y	response matrix (n-by-q)
s1_vec	values for gridsearch (vector of integer valued numerics)
s2_vec	values for gridsearch (vector of integer valued numerics)
pars	list of algorithm parameters; output of set_parameters
seed	set random seed (integer)
k	number of k-folds (integer valued numerics greater than 1)
reps	number of replications (positive integer valued numeric)

Value

A list containing the mean and sd of the error over the replicates as well as the means and standard deviations of the errors across each fold.

Note

See also squared_error, k_folds, tsmvr_solve, tsmvr_cv, tsmvr_replicate, and set_parameters.

spcorum/tsmvr-saved documentation built on Nov. 5, 2019, 9:15 a.m.