[R-sig-ME] Quadratic with Random Offset in One Dimension

[Christopher D. Long]

Hi,

I'm looking to fit a family of quadratics in (x,y) with a random factor
offset in one variable. The model would look like this:

outcome ~ x^2 + x*(y+F) + (y+F)^2 + 1

with F a random factor.

If this were linear in x,y it'd be no problem:

outcome ~ x + y + 1|F.

Is there a way to get either lme4 to estimate a model like this?
If not, what's my best route?
-- 
Christopher D. Long, San Diego Padres, 100 Park Blvd, San Diego CA

"Tick, clong, tick, clong, tick, clong, went the night." - Thurber