randomFillAndRepeatDataRow: Fill unmatched datarow in GLoMo

In GLoMo: Naive General Location Model

Description

Is only called from within predict.GLoMo to handle the unlikely case of a row that has no matching row(s) in the GLoMo uids.

Note: this really only is unlikely in the setting that I plan to use it in, i.e.: impute in the original data from a GLoMo fit to it.

Usage

randomFillAndRepeatDataRow(currow, obsneeded, levelslist, newdata)

Arguments

currow	(unmatched) row of data for which to sample
obsneeded	how many rows should be 'sampled'
levelslist	list holding (possible) levels for all columns in currow
newdata	complete dataset from which currow was one.

Value

a data.frame or numdfr (dependent on the dfr that was used in the original call to GLoMo that holds the sampled data.

Note

Really only intended as an internal function. Kept public for who knows what.

Author(s)

Nick Sabbe ([e-mail])

See Also

GLoMo-package, NumDfr, predict.GLoMo

Examples

iris.md<-randomNA(iris, 0.1)
iris.md.nd<-numdfr(iris.md)
mdrow<-min(which(apply(iris.md, 1, function(currow){any(is.na(currow))})))
randomFillAndRepeatDataRow(currow=iris.md.nd[mdrow,], obsneeded=10,
	levelslist=allLevels(iris.md.nd), iris.md.nd)

GLoMo documentation built on May 2, 2019, 5:26 p.m.