Error in mitml::jomoImpute: Target variables do not contain any missing data.

[stefvanbuuren]

In connection to #378 I came across the following error

suppressPackageStartupMessages(library(mice))
mice(nhanes, 
     blocks = list(c("bmi", "hyp"), "chl"), 
     defaultMethod = c("jomoImpute", "logreg", "polyreg", "polr"))
#> 
#>  iter imp variable
#>   1   1  bmi hyp  chl
#> Error in mitml::jomoImpute(data = data, formula = formula, type = type, : 
Target variables do not contain any missing data.

^{Created on 2021-04-02 by the reprex package (v1.0.0)}

The error occurs when imputing chl. It is not quite clear why jomoImpute() says that this variable has no missing data.

[prockenschaub]

I just stumbled across this in one of my projects. The problem appears to be the type argument passed to mitml::jomoImpute. In mice::sampler line 47 type is set to the row of the predictor matrix corresponding to the block (and therefore a vector of 0/1).

However, ?mitml::jomoImpute states the following

The type interface is designed to provide quick-and-easy imputations using jomo. 
The type argument must be an integer vector denoting the role of each variable in 
the imputation model:

1: target variables containing missing data

2: predictors with fixed effect on all targets (completely observed)

3: predictors with random effect on all targets (completely observed)

-1: grouping variable within which the imputation is run separately

-2: cluster indicator variable

0: variables not featured in the model

A quick and hacky fix that avoids that problem is therefore running

type <- 2 * type
type[blocks[[h]]] <- 1

within the mice:::sampler before the call to mice.impute.jomoImpute but ideally this could be addressed within mice.impute.jomoImpute itself.