Skip to contents

This functions uses posterior samples to quantify uncertainty in the expected measurements from fitted models.

Usage

getExpectedMeasIntervals(expectedMeas, mass = 0.95, method = c("eti", "hdi"))

Arguments

expectedMeas

A data.table of expected measurements for posterior samples, as returned by getExpectedMeas().

mass

Number between 0 and 1 indicating the probability mass for which to calculate the intervals.

method

String indicating the type of interval: 'eti' for equal-tailed using stats::quantile(), or 'hdi' for highest density using HDInterval::hdi().

Value

A data.table containing lower and upper bounds of the expected measurement for each combination of feature, time, and possibly condition and covariate.

Examples

library('data.table')

y = GSE34018$y
metadata = GSE34018$metadata

fit = getModelFit(y, metadata)
fit = getPosteriorFit(fit)
fit = getPosteriorSamples(fit, nPosteriorSamples = 10L)

measFitSamps = getExpectedMeas(
  fit, times = seq(0, 24, 0.5), fitType = 'posterior_samples',
  features = c('13170', '12686'))
measFitInts = getExpectedMeasIntervals(measFitSamps)