r - The ways to construct columns associated with interaction terms in data frame -
i have data frame 6 columns,
dat<-data.frame(x1,x2,x3,x4,x5,x6)
right now, need build 2 columns associated interaction terms, x1*x2
, x3*x4*x5
how in r. there special consideration when of them, such x2
categorical?
i guess function model.matrix
want.
for instance, can fit linear model including variables , interaction terms you're interested in , extract model matrix fitted object
model.matrix(lm(drat ~ mpg * cyl + disp * hp * wt, data = mtcars))
factors need explicitly coded factors, find example below
mtcars$cyl <- factor(mtcars$cyl) model.matrix(lm(drat ~ mpg * cyl + disp * hp * wt, data = mtcars))
the default kind of contrasts used factors treatment coding. can change sum coding (or other codings: ?contr.sum
) using command below
contrasts(mtcars$cyl) <- contr.sum
Comments
Post a Comment