Step: ROC Analysis

Perform ROC analysis using glystats::gly_roc(), extract top 10 variables with highest AUC, and plot ROC curves for these variables using glyvis::plot_roc().

This step requires exp (experiment data).

Usage

step_roc(pos_class = NULL, plot_width = 5, plot_height = 5)

Arguments

pos_class

A character string specifying which group level should be treated as the positive class. If NULL (default), the second level (alphabetically) will be used as the positive class.

plot_width

Width of the plot in inches. Default is 5.

plot_height

Height of the plot in inches. Default is 5.

Value

A glysmith_step object.

Details

Data required:

• exp: The experiment to perform ROC analysis on

Tables generated:

• roc_auc: A table containing the ROC AUC values for all variables

Plots generated:

• roc_curves: ROC curves for the top 10 variables

See also

glystats::gly_roc(), glyvis::plot_roc()

Examples

step_roc()
#> <step "step_roc()"> ROC analysis