ciBA: Bayesian method of CI estimation with different or same...

In proportion: Inference on Single Binomial Proportion and Bayesian Computations

Description

Bayesian method of CI estimation with different or same parameteric values for Beta prior distribution

Usage

ciBA(n, alp, a, b)

Arguments

n	- Number of trials
alp	- Alpha value (significance level required)
a	- Shape parameter 1 for prior Beta distribution in Bayesian model. Can also be a vector of length n+1 priors.
b	- Shape parameter 2 for prior Beta distribution in Bayesian model. Can also be a vector of length n+1 priors.

Details

Highest Probability Density (HPD) and two tailed intervals are provided for all xi = 0, 1, 2 ..n based on the conjugate prior β(ai, bi) (i = 1, 2..n+1) for the probability of success p of the binomial distribution so that the posterior is β(xi + ai, n - xi + bi).

Value

A dataframe with

x	- Number of successes (positive samples)
pomean	- Posterior mean
LBAQ	- Lower limits of Quantile based intervals
UBAQ	- Upper limits of Quantile based intervals
LBAH	- Lower limits of HPD intervals
UBAH	- Upper limits of HPD intervals

References

[1] 2002 Gelman A, Carlin JB, Stern HS and Dunson DB Bayesian Data Analysis, Chapman & Hall/CRC [2] 2006 Ghosh M, Delampady M and Samanta T. An introduction to Bayesian analysis: Theory and Methods. Springer, New York

See Also

prop.test and binom.test for equivalent base Stats R functionality, binom.confint provides similar functionality for 11 methods, wald2ci which provides multiple functions for CI calculation , binom.blaker.limits which calculates Blaker CI which is not covered here and propCI which provides similar functionality.

Other Basic methods of CI estimation: PlotciAS, PlotciAllg, PlotciAll, PlotciBA, PlotciEX, PlotciLR, PlotciLT, PlotciSC, PlotciTW, PlotciWD, ciAS, ciAll, ciEX, ciLR, ciLT, ciSC, ciTW, ciWD

Examples

n=5; alp=0.05; a=0.5;b=0.5;
ciBA(n,alp,a,b)
n=5; alp=0.05; a=c(0.5,2,1,1,2,0.5);b=c(0.5,2,1,1,2,0.5)
ciBA(n,alp,a,b)

proportion documentation built on May 1, 2019, 7:54 p.m.