generateblankdata: Generating a blank data frame

In TiStat/Imputegamlss: Imputing censored covariates with distributional regression - gamlss

Description

This function generates a blank data.frame with uniformly distributed variables specified in varnames. The variables are drawn i.i.d. uniform distribution if correlation = NULL. Otherwise, the data are drawn according to the algorithm:
https://www.r-bloggers.com/easily-generate-correlated-variables-from-any-distribution-without-copulas/
The idea is: covariates specified in variablenames are inversely generated with a surrogate Multivariate normal distribution to establish correlation between mvnormal draws. The draws' cumulative normal probabilities are evaluated in the quantile function of the uniform distribution to arrive at correlated covariates on the interval [0,1]. However, they are not uniformly distributed, as the Blogpost propagates.

Usage

generateblankdata(varnames, n, correlation = NULL)

Arguments

varnames	character vector. Specifies variables to be created.
n	integer. Number of observations to be generated.
correlation	Symmetric correlation matix.

Value

data.frame

TiStat/Imputegamlss documentation built on May 20, 2019, 9:25 a.m.