nma.prior: Determine the appropriate value to be used for default priors

In augustinewigle/StanNet: StanNet: A Variant of BUGSnet Using Stan

Description

Finds the largest maximum likelihood estimator from individual trials as in van Valkenhoef, G et al. 2012. This function was originally created in BUGSnet (https://github.com/audrey-b/BUGSnet) without documentation or comments. This version has small changes to avoid warnings during the CRAN check.

Usage

nma.prior(data.nma, outcome, scale, N, sd = NULL, time = NULL)

Arguments

data.nma	A data object created by running data.prep()
outcome	A string for the column name containing the outcome
scale	A string indicating the scale on which the relative treatment effects are being measures
N	A string for the column name containing the sample sizes
sd	A string for the column name containing the standard errors (only required for continuous outcomes)
time	A string for the column names containing the follow-up times or person-time at risk for rate or rate2 data

Value

The largest MLE to be used in determining default priors for mu, d and sigma

augustinewigle/StanNet documentation built on July 21, 2020, 12:13 a.m.