Heatmap of the results of cross-validation for Bhat obtained from the rrda.cv function.

This function creates a heatmap to visualize the Mean Squared Error (MSE) results from the cross-validation of the Bhat matrix obtained from the rrda.cv function. The heatmap displays the MSE for different ranks of Bhat and values of the regularization parameter lambda, allowing users to visually assess the best combination of rank and lambda. The function also allows the user to highlight the points corresponding to the minimum MSE and the 1-standard error rule, helping to identify optimal model parameters.

Usage

rrda.heatmap(
  cv_result,
  nrank = NULL,
  min_l = NULL,
  max_l = NULL,
  highlight_min = TRUE,
  title = NULL
)

Arguments

cv_result

A result list from the function rrda.cv, containing a matrix of MSE values for each rank and lambda, and a vector of lambda values.

nrank

A numeric vector specifying the ranks of Bhat to be plotted. Default is NULL, which plots all ranks.

min_l

Minimum lambda value to be plotted. Default is NULL, which uses the minimum lambda value in cv_result.

max_l

Maximum lambda value to be plotted. Default is NULL, which uses the maximum lambda value in cv_result.

highlight_min

Logical indicating if the marks should be plotted on the best prediction point, and 1se point. Default is TRUE.

title

Title of the figure

Value

A heatmap of MSE cross-validation results.

Examples

if (FALSE) { # \dontrun{
set.seed(10)
simdata<-rdasim1(n = 100,p = 200,q = 200,k = 3) # data generation
X <- simdata$X
Y <- simdata$Y

cv_result<- rrda.cv(Y = Y, X = X, maxrank = 5, nfold = 5) # cv
rrda.summary(cv_result = cv_result)
rrda.heatmap(cv_result=cv_result)
} # }