boot_auc_folds: Compute bootstraps of the mean ROC
In mattdneal/FAIMSToolkit: Tools to aid in the analysis of FAIMS data

Compute the ROC curve separately for each fold, and compute the mean ROC curve, instead of pooling predictions.

1
2
3

boot_auc_folds(predictions, folds, targets, num_samples = 1000,
  alt_predictions = NULL, fixed_axis = c("sensitivity", "specificity"),
  points = seq(0, 1, by = 0.01), stratified = T)

`predictions`	sample predictions
`folds`	vector of fold classification for each sample
`targets`	target values
`num_samples`	number of bootstrap samples to take
`alt_predictions`	alternative predictions to compare
`fixed_axis`	whether to fix sensitivity or specificity in the ROC curves
`points`	number of points to calculate for ROC curves