[R] a-priori orthogonal contrasts

[Keith Alan Chamberlain]

Dear R-helpers,

I have some stacked data with which I want to generate orthogonal contrasts - with the caveat that I determine what means comparisons are taking place a-priori.

Does it matter for e.g. 'contr.SUM' that the data are stacked, and does 'contr.SUM' allow me to set the contrasts efficiently even if the data are unballanced? Is there an alternative in R I should consider?

Sincerely,
KeithC.
Psych Undergrad, CU Boulder (US)