mice.mids {mice} | R Documentation |
Multivariate Imputation by Chained Equations (Iteration Step)
Description
Takes a mids
object, performs maxit
iterations and
produces a new object of class "mids"
.
Usage
mice.mids(obj, newdata = NULL, maxit = 1, printFlag = TRUE, ...)
Arguments
obj |
An object of class |
newdata |
An optional |
maxit |
The number of additional Gibbs sampling iterations. The default is 1. |
printFlag |
A Boolean flag. If |
... |
Named arguments that are passed down to the univariate imputation functions. |
Details
This function enables the user to split up the computations of the Gibbs sampler into smaller parts. This is useful for the following reasons:
To add a few extra iteration to an existing solution.
If RAM memory is exhausted. Returning to prompt/session level may alleviate such problems.
To customize convergence statistics at specific points, e.g., after every
maxit
iterations to monitor convergence.
The imputation model itself is specified in the mice()
function
and cannot be changed in mice.mids()
. The state of the random
generator is saved with the mids
object. This ensures that the
imputations are reproducible.
Value
mice.mids
returns an object of class "mids"
.
See Also
complete
, mice
, set.seed
,
mids
Examples
imp1 <- mice(nhanes, maxit = 1, seed = 123)
imp2 <- mice.mids(imp1)
# yields the same result as
imp <- mice(nhanes, maxit = 2, seed = 123)
# verification
identical(imp$imp, imp2$imp)
#