generate.confidence.intervals: Confidence interval function

In itslwg/superlearnerr: Package for the SuperLearner vs clincians study

Description

This function generates confidence intervals around difference of two given point estimates using empirical bootstrapping.

Usage

generate.confidence.intervals(study_sample, model_or_pointestimate, func,
  samples, diffci_or_ci, outcome_name)

Arguments

study_sample	The study sample list. No default.
model_or_pointestimate	Character vector describing predictions or point estimates to analyse. For example, pred_cat and clinicians_predictions predictions for model.review.AUROCC or NRI+ and NRI- point estimates for model.review.reclassification. No default.
func	Function that generates key statistic. For example, model.review.AUROCC that generates AUROCC of a given model, or model.review.reclassification that generates reclassification elements. No default.
samples	Samples as prepared with train.predict.bssamples. No default.
diffci_or_ci	String. Whether to return confidence interval on difference of model_or_pointestimates, or return confidence intervals on model_or_pointestimates separately or no confidence intervals. Must be one of c("diff", "ci", "none")
outcome_name	Name of outcome variable. No default.

itslwg/superlearnerr documentation built on July 31, 2020, 7:37 a.m.