GNGTools (version 1.0.0)

return.ts.ng.int.req.df: Return two-sample normal-gamma predictive probability

Description

Return two-sample normal-gamma predictive probability

Usage

return.ts.ng.int.req.df(
  mu.0.t = 0,
  n.0.t = 1e-04,
  alpha.0.t = 0.25,
  beta.0.t = 1,
  mu.0.c = 0,
  n.0.c = 1e-04,
  alpha.0.c = 0.25,
  beta.0.c = 1,
  xbar.t = c(1.9, 2, 2.1, 2.05),
  s.t = c(2, 2.1, 1.9, 2.04),
  n.t = c(10, 20, 30, 40),
  xbar.c = c(1, 1.1, 1.5, 1.25),
  s.c = c(1.9, 2, 2.5, 2.25),
  n.c = c(10, 20, 30, 40),
  Delta.lrv = 1.25,
  Delta.tv = 1.75,
  tau.tv = 0.1,
  tau.lrv = 0.8,
  tau.ng = 0.65,
  xbar_ng = NULL,
  xbar_go = NULL,
  go.thresh = 0.8,
  ng.thresh = 0.8,
  n.MC = 1000
)

Value

A dataframe is returned

Arguments

mu.0.t

prior mean for treatment group

n.0.t

prior effective sample size parameter for treatment group

alpha.0.t

prior alpha parameter for treatment group

beta.0.t

prior beta parameter for treatment group

mu.0.c

prior mean for control group

n.0.c

prior effective sample size parameter for control group

alpha.0.c

prior alpha parameter for control group

beta.0.c

prior beta parameter for control group

xbar.t

treatment mean

s.t

treatment sd

n.t

treatment sample size

xbar.c

control mean

s.c

control sd

n.c

control sample size

Delta.lrv

TPP Lower Reference Value aka Min TPP

Delta.tv

TPP Target Value aka Base TPP

tau.tv

threshold associated with Base TPP

tau.lrv

threshold associated with Min TPP

tau.ng

threshold associated with No-Go

xbar_ng

Leave NULL to determine what is required or supply a value

xbar_go

Leave NULL to determine what is required or supply a value

go.thresh

If the predictive probability that study will conclude as 'Go' is larger than this threshold: Declare 'Interim go'.

ng.thresh

If the predictive probability that study will conclude as 'No-Go' is larger than this threshold: Declare 'Interim no-go'

n.MC

Monte Carlo simulation size

Examples

holdit <- return.ts.ng.int.req.df()
head(holdit)