0

Trying to use gam::step.gam to build a model. Don't understand why that function asks for a formula if the input model object already has an embedded formula (from when it was fit via gam::gam)

Why ask for the formula twice? When would I want to use different formulas between the two functions?

My review of help("gam") and help("step.gam") didn't uncover anything to explain why I'd use one formula when first fitting model with gam and another when feeding that model to step.gam.

MCornejo
  • 327
  • 1
  • 12
  • 1
    Have you studied `help("step.gam")`? – Roland Jan 05 '17 at 14:29
  • 1
    Welcome to Cross Validated. Questions that are solely about programming is [off-topic](http://stats.stackexchange.com/help/on-topic) for this site and may be closed. If you have a statistical question, please edit your post to emphasize it. Programming questions may be better received at StackOverflow (do *not* cross-post there, ask the moderators to migrate it) or you can browse the [list of internet support for statistics software](http://meta.stats.stackexchange.com/questions/793/internet-support-for-statistics-software) that we maintain. – Marquis de Carabas Jan 05 '17 at 15:24
  • 2
    This is not a question about how GAMs work, but how the `R` functions work. This is a question about using `R` code, not about statistics. If you want to know why the package developers designed the functions the way they did, you should ask them. This is off topic here. – gung - Reinstate Monica Jan 05 '17 at 15:47
  • Migrating to SO per request of OP. – whuber Jan 05 '17 at 16:28

0 Answers0