cor_mi_marginplot: MI Marginplots of all highly correlated variables

In mmaechler/miP: Multiple Imputation Plots

Description

Computes correlations between all variables taking all observations without missing values on both variables. Then plots MI Marginplots of all highly correlated variables.

Usage

cor_mi_marginplot(input, cor = 0.5, obs = 0.5)

Arguments

input	MI object produced by mi, mice or Amelia.
cor	MI Marginplots between two variables are drawn only if their correlation is higher than the absolute value of this parameter.
obs	MI Marginplots between two variables are drawn only if the rate of observed values is higher than this parameter.

Details

This plot is inspired by the marginplot function in R package VIM. For details see marginplot there.

Value

Returns a correlation matrix, all correlations lower than abs(cor) are NA.

Author(s)

Paul Brix

References

R package "VIM: Visualization and Imputation of Missing Values" by Matthias Templ, Andreas Alfons, Alexander Kowarik. cran.r-project.org/package=VIM

See Also

mi_marginplot and marginplot from VIM.

Examples

data(FLAS)
require("mice")
FLASmi <- mice(FLAS)
cor_mi_marginplot(FLASmi)

mmaechler/miP documentation built on May 23, 2019, 5 a.m.