multinomial.conf: Simultaneous Confidence Functions for Multinomial...

In ddarmon/clp: R Package for Computing Confidence Functions to Accompany CLP

Description

Confidence functions for a multinomial proportions based on the mid P-value for each proportion, with balancing across proportions via bootstrapping with Beran's simultaneous confidence sets.

Usage

multinomial.conf(N, plot = TRUE, conf.level = 0.95, B = 2000, col = NULL)

Arguments

N	a vector of counts in each category from a sample of size n
plot	whether to plot the confidence density and curve
conf.level	the confidence level for the confidence interval indicated on the confidence curve
B	the number of bootstrap samples to use with Beran's method
col	a vector of colors to use for plotting confidence functions of each proportion

Value

A list of lists of lists containing the confidence functions pconf, dconf, cconf, and qconf for each proportion without correction for multiple comparison (singlecomp) or with correction for multiple comparison (multicomp).

References

Tore Schweder and Nils Lid Hjort. Confidence, Likelihood, Probability. Vol. 41. Cambridge University Press, 2016.

Tore Schweder. "Confidence nets for curves." Advances In Statistical Modeling And Inference: Essays in Honor of Kjell A Doksum. 2007. 593-609.

Rudolf Beran. "Balanced simultaneous confidence sets." Journal of the American Statistical Association 83.403 (1988): 679-686.

Examples

# A hypothetical Punnett square experiment with peas.

N <- c(315, 108, 101, 32)

names(N) <- c('Round, Yellow', 'Round, Green', 'Wrinkled, Yellow', 'Wrinkled, Green')

col <- c('darkgoldenrod', 'darkgreen', 'darkgoldenrod1', 'darkolivegreen')

peas.conf <- multinomial.conf(N, col = col)

# Confidence intervals without and with correction for multiple comparisons:
peas.conf$singlecomp$`Round, Yellow`$qconf(c(0.025, 0.975))
peas.conf$multicomp$`Round, Yellow`$qconf(c(0.025, 0.975))

ddarmon/clp documentation built on Jan. 25, 2021, 6:22 p.m.