I'm performing numereous loess analyses. The problem now is that I have
to write for each model the formula. Is there a way to do this as in
the parametric regression where one can automatically obtain the nbest models
such as in stepwise?
If not has anybody written code which could do the following?
y is the dependent variable
x1, x2, x3, x4 and x5 the independent variables.
I would now like to obtain the summary.loess of each permutation of
x1 to x5 under the condition that always 3 independent variables enter the
model.
y~x1+x2+x3
y~x1+x2+x4
y~x1+x2+x5
y~x1+x3+x4
y~x1+x3+x5
y~x1+x4+x5
y~x2+x3+x4
y~x2+x3+x5
y~x2+x4+x5
y~x3+x4+x5
Well, finally this function should be rather flexible, though that more than
5 independent variables can be used and that each regression model takes
any amount of independent variables (at least four).
Thank you in advance
Greetings
Bernhard