[R] bestglm - number of variables

pieniu petersmith at op.pl
Tue Jul 17 12:06:44 CEST 2012


Hello,

I have a question concerning the "bestglm" procedure (function that finds
the best subset of explanatory variables for the variable we want to model
by checking all possible models), available in the package "bestglm".
When I use the function for a database that has over 45 variables (that is
44 potential explanatory variables and one that will be explained by the
model) the algorithm is suspended (it looks like it works but for a very
long period of time nothing happens and eventually one has to quit the whole
programme as it does not respond). With fewer than 45 variables the
procedure runs quite fast.
I tried to limit the maximum size of models that the function searches
through (that is the number of explanatory variables within the model) by
the parameter "nvmax", without limiting the whole set of potential
explanatory variables, but the problem did not disappear.
I would be grateful for any hints on this topic.

Regards,

Piotr Piekos,
Warsaw School of Economics

--
View this message in context: http://r.789695.n4.nabble.com/bestglm-number-of-variables-tp4636736.html
Sent from the R help mailing list archive at Nabble.com.



More information about the R-help mailing list