Monotone data imputation requires changing the predictor matrix

[stefvanbuuren]

Jonathan Bartlett alerted me to an error in the example in https://stefvanbuuren.name/fimd/sec-monotone.html. One-step convergence in monotone data imputation requires three changes to the default arguments in mice:

1. Set visitSequence = "monotone"
2. Set maxit = 1
3. Remove later variables from the predictorMatrix.

The sample code in the book specifies 1 and 2, but not 3. If we do not do 3, then convergence within 1 iteration is not guaranteed.

Probably most convenient fix is to let mice automatically change the predictor matrix if the user specifies visitSequence = "monotone".

[stefvanbuuren]

Commit 8f78a27 adds a function edit.predictorMatrix() that changes predictorMatrix to a monotone pattern if visitSequence = "monotone" and if maxit = 1.

Adds to documentation:
Special case: If you specify both visitSequence = "monotone" and maxit = 1, then the procedure will edit the predictorMatrix to conform to the monotone pattern. Realize that convergence in one iteration is only guaranteed if the missing data pattern is actually monotone. The procedure does not check this.