[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: creating interaction variables
At 02:45 AM 6/12/2004 +0900, Joseph Coveney wrote:
Richard is absolutely right and I stand corrected. The interaction term
I stand corrected too because I have -xi3- and didn't realize it could do
this! Also, it includes the lower order terms along with the interaction,
which is good, e.g.
command that allows for continuous-by-continuous interaction terms, in
to those involving categorical-predictor variables, is -xi3- from Michael
Mitchell and Phil Ender. It's an extension of -xi- that is downloadable from
SSC or from the ULCA site. -help xi3- (after installing the package) will
showTewodaj the way to the goal.
xi3: reg y x1*x2
results in y being regressed on x1, x2, and x1*x2.
-xi3- has a lot of other nice options too.
* For searches and help try: