linear_hypothesis: Test Linear Hypotheses for Posterior Samples

In donaldRwilliams/blsmeta: Bayesian Location-Scale Meta-Analysis

Description

Test Linear Hypotheses for Posterior Samples

Usage

linear_hypothesis(
  lin_comb,
  obj,
  cred = 0.95,
  sub_model = NULL,
  print_names = FALSE,
  rope = NULL,
  contrast = NULL
)

Arguments

lin_comb	A string specifying a linear combination of variables, or a list of variable names if using contrast.
obj	An object of class blsmeta
cred	The level for which a credible interval should be computed.
sub_model	character. Which sub-model for the hypotheses, with options including location and scale. The defaul is NULL, so either is required.
print_names	logical. Should the parameter names be printed ? This is useful for specifying the hypotheses, as the names need to be exact.
rope	Specify a ROPE. Optional.
contrast	A contrast matrix specifying which combinations to test. Optional.

Value

An object of class linear_hypothesis

Examples

library(psymetadata)

fit <-  blsmeta(yi = yi, 
                vi = vi, 
                es_id = es_id,
                study_id = study_id,
                data = gnambs2020)
                
linear_hypothesis(obj = fit, 
                  lin_comb = "scale3_Intercept > scale2_Intercept",
                  sub_model = "scale")

donaldRwilliams/blsmeta documentation built on Dec. 20, 2021, 12:12 a.m.