Skip to contents

Plot Cross-Validation Results

Source: R/plot_nmf.R
plot.nmfCrossValidate.Rd

Visualize NMF cross-validation results showing test (and optionally train) loss across candidate ranks. Useful for selecting the optimal factorization rank.

Usage

# S3 method for class 'nmfCrossValidate'
plot(x, show_train = NULL, point_size = 3, interactive = FALSE, ...)

Arguments

x

object of class nmfCrossValidate (a data.frame from nmf(k = 2:10, test_fraction = 0.1))

show_train

logical, overlay train loss on the plot (default TRUE if train_mse column is available and non-NA)

point_size

size of data points (default 3)

interactive

create interactive plotly plot (default FALSE)

...

additional arguments (unused)

Value

ggplot2 or plotly object

See also

nmf

Examples

# \donttest{
library(Matrix)
A <- abs(rsparsematrix(50, 30, 0.3))
cv_result <- nmf(A, k = 2:6, test_fraction = 0.1, cv_seed = 1, maxit = 20)
plot(cv_result)

# }