IntrinsicInference: Intrinsic inference on the rate ratio.

In stla/brr: Bayesian Inference on the Ratio of Two Poisson Rates

Description

Intrinsic inference on the rate ratio.

Usage

intrinsic_phi0(phi0, x, y, S, T, a = 0.5, b = 0, c = 0.5, d = 0,
  beta_range = TRUE, tol = 1e-08, ...)

intrinsic_phi0_sims(phi0, x, y, S, T, a = 0.5, b = 0, c = 0.5, d = 0,
  nsims = 1e+06)

intrinsic_estimate(x, y, S, T, a = 0.5, b = 0, c = 0.5, d = 0,
  otol = 1e-08, ...)

intrinsic_H0(phi.star, alternative, x, y, S, T, a = 0.5, b = 0, c = 0.5,
  d = 0, ...)

intrinsic_bounds(x, y, S, T, a = 0.5, b = 0, c = 0.5, d = 0,
  conf = 0.95, parameter = "phi", otol = 1e-08, ...)

Arguments

phi0	the proxy value of phi
x,y	Observed counts
S,T	sample sizes
a,b,c,d	Prior parameters
beta_range	logical, if TRUE (default), an internal method is used to avoid a possible failure in numerical integration; see the main vignette for details
tol	accuracy requested
...	other arguments passed to integrate
nsims	number of simulations
otol	desired accuracy for optimization
phi.star	the hypothesized value of phi
alternative	alternative hypothesis, "less" for H1: phi0 < phi.star, "greater" for H1: phi0 > phi.star
conf	credibility level
parameter	parameter of interest: relative risk "phi" or vaccine efficacy "VE"

Value

intrinsic_phi0 returns the posterior expected loss, intrinsic_estimate returns the intrinsic estimate, intrinsic_H0 performs intrinsic hypothesis testing, and intrinsic_bounds returns the intrinsic credibility interval.

Examples

a<-0.5; b<-0; c<-1/2; d<-0; S<-100; T<-S; x<-0; y<-20
intrinsic_phi0(0.5, x, y, S, T, a, b, c, d)
intrinsic_phi0_sims(0.5, x, y, S, T, a, b, c, d)
intrinsic_estimate(x, y, S, T, a, b, c, d)
bounds <- intrinsic_bounds(x, y, S, T, a, b, c, d, conf=0.95); bounds
ppost_phi(bounds[2], a, b, c, d, S, T,  x, y)- ppost_phi(bounds[1], a, b, c, d, S, T, x, y)

stla/brr documentation built on May 30, 2019, 5:46 p.m.