miList
whose elements correspond to
multiple imputations.
The generator function for this class of objects is
miList
.
miList
objects are commonly created as the output of
miEval
and
miApply
, functions that perform calculations on multiple imputations.
The
"miList"
class of objects has associated methods:
anyMissing.miList
,
as.miVariable.miList
,
data.frameAux.miList
,
miDiscard.miList
,
miTrim.miList
,
numberMissing.miList
, and
print.miList
.
Other functions should be used within
,
e.g.
miEval(summary(x))
.
An
miList
is a list with class
"miList"
. Each component of
this list corresponds to one set of (multiple) imputation; there must
be at least two imputations.
This is a very general class.
Each component of the list corresponds to a single set of multiple
imputations; the components typically have the same structure
but different numerical values.
For example, if a data frame contains
M
sets of
multiple imputations, then the output of
miEval
for
a linear regression using that data will be an
miList
with
M
components, each of which is an
lm
object obtained by
replacing missing values in the data frame with one of the
M
sets of imputations, then running the regression.
x <- miList(list(3, 4, 5)) x # Default names are 1, 2, ... oldClass(x) # has old-style class "miList"