coef_diff_by_group {breathtestcore}

Tabulates breath test parameter differences of groups

Description

Given a fit to 13C breath test curves, computes between-group confidence intervals and p-values, for examples of the half emptying time t50, with correction for multiple testing.

Usage

coef_diff_by_group(fit, mcp_group = "Tukey", reference_group = NULL, ...)

Arguments

fit

Object of class breathtestfit, for example from nlme_fit, nls_fit

mcp_group

"Tukey" (default) for all pairwise comparisons, "Dunnett" for comparisons relative to the reference group.

reference_group

Used as the first group and as reference group for mcp_group == "Dunnett"

...

Not used

Value

A tibble of class coef_diff_by_group with columns

parameter

Parameter of fit, e.g. beta, k, m, t50

method

Method used to compute parameter. exp_beta refers to primary fit parameters beta, k, m. maes_ghoos uses the method from Maes B D, Ghoos Y F, Rutgeerts P J, Hiele M I, Geypens B and Vantrappen G 1994 Dig. Dis. Sci. 39 S104-6. bluck_coward is the self-correcting method from Bluck L J C and Coward W A 2006

groups

Which pairwise difference, e.g solid - liquid

estimate

Estimate of the difference

conf.low, conf.high

Lower and upper 95 A comparison is significantly different from zero when both estimates have the same sign.

p.value

p-value of the difference against 0, corrected for multiple testing

Examples

library(dplyr)
data("usz_13c")
data = usz_13c %>%
  dplyr::filter( patient_id %in%
    c("norm_001", "norm_002", "norm_003", "norm_004", "pat_001", "pat_002","pat_003")) %>%
  cleanup_data()
fit = nls_fit(data)
coef_diff_by_group(fit)

fit = nlme_fit(data)
coef_diff_by_group(fit)

# TODO: Add example for Stan fit typecast to class \code{breathtestfit} to compute
# confidence intervals instead of credible intervals