Hello, I am working on a variable selection problem and would like to have some suggestions. Thank you. In my data, the number of observations/samples is much less than the number of variables. And I am not interested in generating only a few models, instead I will need a couple of hundred models. For each model, I only need a fixed number of variables, in other word, with a specific cardinality. I've tried leaps(subselect package) and regsubsets(leaps package). However, I have to reduce the number of variables is using leaps in subselect package which is not I want and the regsubsets in leaps package doesn't read a specific cardinality. It accepts a maximal subset size. Thank you. -- View this message in context: http://r.789695.n4.nabble.com/subsets-with-a-small-cardinality-for-variable-selection-tp2965552p2965552.html Sent from the R help mailing list archive at Nabble.com.