Recall that the main requirement of an ldframe object is that
its Subject column be block-sorted, with unique subject codes
appearing in contiguous blocks. It is however possible to select rows
from an ldframe object such that the resultant object does
not comply with this restriction. If you attempt to do so, a warning
will be given and the result object will be of class
data.frame instead of ldframe.
![\begin{Example}
\gt baddf <- data.ldf[c(1:5,10:6),]
Warning messages:
Selection...
....81 1 5
7 2.11 1 4
6 3.25 1 3
\gt class(baddf)
[1] ''data.frame''\end{Example}](img28.gif)
Observe that the observations for Subject 1 no longer appear in a
contiguous block. If such a situation arises, you can convert the
data.frame back into an ldframe object with
as.ldframe, which arranges the Subject column into
contiguous blocks where necessary. By default, as.ldframe
also sorts by Time within each Subject code. This is for
convenience only; to suppress sorting by Time use the option
sort.time=F to as.ldframe:
