convert_r: Convert r for NORTA

In faux: Simulation for Factorial Designs

Convert r for NORTA

Description

Given a target r-value, returns the correlation you need to induce in a bivariate normal distribution to have the target correlation after converting distributions.

Usage

convert_r(
  target_r = 0,
  dist1 = "norm",
  dist2 = "norm",
  params1 = list(),
  params2 = list(),
  tol = 0.01
)

Arguments

target_r	The target correlation
dist1	The target distribution function for variable 1 (e.g., norm, binom, gamma, truncnorm)
dist2	The target distribution function for variable 2
params1	Arguments to pass to the functions for distribution 1
params2	Arguments to pass to the functions for distribution 2
tol	Tolerance for optimise function

Details

See Distributions for distributions and their various arguments to specify in params1 and params2.

Value

r-value to induce in the bivariate normal variables

Examples

convert_r(target_r = 0.5, 
          dist1 = "norm", 
          dist2 = "binom", 
          params1 = list(mean = 100, sd = 10),
          params2 = list(size = 1, prob = 0.5))
          
convert_r(target_r = 0.5, 
          dist1 = "norm", 
          dist2 = "likert", 
          params1 = list(mean = 100, sd = 10),
          params2 = list(prob = c(5, 10, 20, 30, 20)))

faux documentation built on April 20, 2023, 9:13 a.m.