[R] model diagnostics for MatrixModels

[Carson Farmer]

Dear list,

I have been working with the MatrixModels package quite a bit this
week, and it is proving to be extremely valuable for my current work
(I am working with several factors with many levels, leading to a
sparse model matrix). However, as my knowledge of statistical theory
leaves much to be desired, there are certain aspects of model
evaluation etc that I am having trouble with. Has anyone developed any
examples of model diagnostics using the glpModel class (from the
MatrixModels package)? Something akin to 'summary', or perhaps helper
functions such as 'logLik', 'AIC', 'predict' or even a means of
running some of the tests from package 'lmtest'? The structure of the
'glpModel' class is pretty straight-forward, so I am able to extract
*some* of the relevant information to perform *some* of these tests
myself, however, an example of performing some 'standard' model
diagnostics or tests/comparisons would be very helpful (In this case I
am really only interested in tests etc relevant to GLMs). I'm not
against writing some of these functions myself, so any
tips/suggestions/resources/examples are appreciated. Furthermore, does
anyone know if there facilities to obtain, for example, the rank of
the (sparse) model matrix efficiently? Quite a few tests require rank,
but computing this separately using for example rankMatrix from
package Matrix uses up tonnes of memory, which is exactly what I was
trying to avoid by using sparse matrices and glm4 :-p

Thanks for any pointers,

Carson

> sessionInfo()
R version 2.12.2 (2011-02-25)
Platform: x86_64-pc-linux-gnu (64-bit)

locale:
 [1] LC_CTYPE=en_IE.utf8       LC_NUMERIC=C
 [3] LC_TIME=en_IE.utf8        LC_COLLATE=en_IE.utf8
 [5] LC_MONETARY=C             LC_MESSAGES=en_IE.utf8
 [7] LC_PAPER=en_IE.utf8       LC_NAME=C
 [9] LC_ADDRESS=C              LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_IE.utf8 LC_IDENTIFICATION=C

attached base packages:
[1] splines   stats     graphics  grDevices utils     datasets  methods
[8] base

other attached packages:
 [1] MatrixModels_0.2-1 pscl_1.03.6        gam_1.03           akima_0.5-4
 [5] coda_0.13-5        mvtnorm_0.9-92     lmtest_0.9-27      zoo_1.6-4
 [9] Matrix_0.999375-48 lattice_0.19-17    MASS_7.3-7

loaded via a namespace (and not attached):
[1] grid_2.12.2  tools_2.12.2


-- 
Carson J. Q. Farmer
ISSP Doctoral Fellow
National Centre for Geocomputation
National University of Ireland, Maynooth,
http://www.carsonfarmer.com/