search for: scrunge

...sometimes gets re-arranged compared to the original fit, and then disaster results. Specifically, the 'variables' and 'predvars' attributes of a 'terms' object get out of synch. This only happens when the terms in the original formula get re-ordered during fitting: test> scrunge.data_ data.frame( contin=1:10, discrete=factor( rep( c( 'cat', 'dog'), 5)), resp=runif( 10)) test> lm.ok_ lm( resp ~ discrete + contin %in% discrete, data=scrunge.data) test> predict( lm.ok, scrunge.data) # no problemo 1 2 3 4 5...

...is very similar to code that lives inside 'tools:::.install_package_Rd_objects'. The problems occur with just-released R2.10.0 on Windows. The examples below use files which can be found at ftp://ftp.csiro.au/MarkBravington, but you'll obviously need to modify the paths. The file "scrunge.Rd" is just "Rdiff.Rd" from the 'tools' package. The file "fakepack.7z" should unzip to create a fake package with a DESCRIPTION file and a "man" directory that contains two Rd files. Transcript of first example: eglist <- list( scrunge=tools:::prepare...

...t new factor levels, even if the "new" levels are merely levels in the original that didn't occur in the original fit and were sensibly dropped, and that don't occur in the prediction data either. (At least if `drop.unused.levels' was set to TRUE, which the default.) test> scrunge.data.2_ data.frame( y=runif( 3), disc=factor( c( 'cat', 'dog', 'cat'), levels=c( 'cat', 'dog', 'earwig'))) test> lm.predbug.2_ lm( y~disc, data=scrunge.data.2) test> predict(lm.predbug.2, newdata=scrunge.data.2) Error in model.frame.default(obje...