README.md

In bojinov/gcImp: Generate multiple imputations using a Gaussian copula approach

gcImp

An R package for generating multiple imputations using a Gaussian copula This package generates multiple imputations using Peter Hoff's sbgcop package. The package can also convert the output to allow for seamless integration with the mice package, which can be used to analyze the multiple imputed data sets.

```r

Install using devtools

library(devtools) devtools::install_github("bojinov/gcImp") Here is an exmaple:

Generate N samples from a multivariate normal

N <- 200 rho <- matrix(0.3, 2, 2) diag(rho) <- 1

Compute the Choleski decomposition of rho.

rho.chol <- chol(rho) samples.mvn <- matrix(rnorm(N * 2), ncol = 2) %*% rho.chol

Delete some of the values

p <- rep(0.2, N) # MCAR

p <- 1/(1 + exp(0.2 - samples.mvn[, 1])) #MAR

R <- sapply(1:N, function(jj) sample(c(NA, 1), size = 1, prob = c(p[jj], 1 - p[jj])))

Generate the observed data

samples.mvn[, 2] <- samples.mvn[, 2] * R out <- gcImp(samples.mvn) print(out)

Check the trace plot of the latent correlation

plot(out$sbgcop.out$C.psamp[1,2,], type = "l")

Check the trace plot of the mean of the imputed data

plot(colMeans(out$sbgcop.out$Y.imput[,2,]), type = "l")

Not run:

Further checks can be performed using the mcmcplots package

library(mcmcplots) mcmcplots::mcmcplot(colMeans(out$sbgcop.out$Y.imput[,2,])) mcmcplots::mcmcplot(out$sbgcop.out$C.psamp[1,2,])

End(Not run)

Convert to a class "mira" and use MICE for analysis

imp <- gc.as.mids(out)

Stack the imputations and run a linear regression

stacked <- mice::complete(imp, "long") fit <- lm(V1 ~ V2, data = stacked) coef(fit)

Fir separate regressions and combine the output using Rubin's rules

fit <- with(imp, lm(V1 ~ V2)) est <- mice::pool(fit) summary(est)

Plot the imputed against the observed

mice::densityplot(imp)

bojinov/gcImp documentation built on May 29, 2019, 10:35 a.m.