alternatives: Generate realizations of beta alternative distributions

In PoolBal: Balancing Central and Marginal Rejection of Pooled p-Values

Generate realizations of beta alternative distributions

Description

These functions can be used to generate samples of p-values all following a beta distribution (H4) or following either uniform or beta distributions according to proportion eta (H3).

Usage

rBetaH4(a, b = 1/w + a * (1 - 1/w), w = (1 - a)/(b - a), M = 2, N = 10)

rBetaH3(
  a,
  b = 1/w + a * (1 - 1/w),
  w = (1 - a)/(b - a),
  eta = 0.5,
  M = 2,
  N = 10
)

Arguments

a	first beta parameter, numeric between 0 and infinity
b	second beta parameter, numeric between 0 and infinity
w	UMP parameter between 0 and 1
M	number of p-values per realization
N	number of realizations
eta	numeric between 0 and 1, proportion of non-null tests per sample

Details

Alternatives

These functions are provided as a convenience, and support a/b (shape1/shape2) or a/w specification of beta parameters.

Value

An N by M matrix of simulated p-values.

Functions

• rBetaH4(): iid Beta(a,w) p-values

• rBetaH3(): M*eta iid Beta(a,w) p-values, others uniform

Author(s)

Chris Salahub

Examples

rBetaH4(a = 0.5, b = 1.5, M = 10, N = 100)
rBetaH3(a = 0.5, b = 1.5, eta = 0.5, M = 10, N = 100)

PoolBal documentation built on Nov. 22, 2023, 5:07 p.m.