Skip to content

How to iterate through different arguments?

6 messages · lord12, Thomas Stewart, Gabor Grothendieck

Original
lord12
# 
If I have a model line = lm(y~x1) and I want to use a for loop to change the
number of explanatory variables, how would I do this?

So for example I want to store the model objects in a list.

model1 = lm(y~x1)
model2 = lm(y~x1+x2)
model3 = lm(y~x1+x2+x3)
model4 = lm(y~x1+x2+x3+x4)
model5 = lm(y~x1+x2+x3+x4+x5)...
model10. 

model_function = function(x){
for(i in 1:x) {
}
If x =1, then the list will only add model1. If x =2, then the list will add
both model1 and model2. If x=3, then the list will add model1 model 2 and
model3 and so on. How do I translate this into code?
Gabor Grothendieck
#
On Sun, Oct 3, 2010 at 4:29 PM, lord12 <trexinsd at yahoo.com> wrote:
Here are a couple of approaches.  The first one is simpler and may be
adequate.  The second has the advantage that it writes out the formula
in full, fo, which is shown in the output:

lapply(1:4, function(i) lm(y1 ~., anscombe[c(1:i, 5)]))

lapply(1:4, function(i) {
	fo <- formula(model.frame(y1 ~., anscombe[c(1:i, 5)]))
	do.call("lm", list(fo, quote(anscombe)))
})
2 days later
lord12
# 
Thanks a lot! I have one other question,however. If I want to have two
garchFits,  garchFit(~aparch(1,1), data = sunspot.year) and
garchFit(~arma(2,1)+apaarch(1,1), data = sunspot.year), how do I automate
the creation of these models? 

for(k in 1:2)
{
  garchFit~...,data = sunspot.year), 
}

I am not sure what to put in the ...
Gabor Grothendieck
#
On Wed, Oct 6, 2010 at 10:16 AM, lord12 <trexinsd at yahoo.com> wrote:
Set up a character vector of terms, Terms, such as

  Terms <- c("x1", "x2", "x3", "x4")

and then replace the fo<- line in my second solution with:

  fo <- as.formula(paste("y1", paste(Terms[1:k], collapse = "+"), sep = "~"))