- Default method is Nelder-Mead (no gradients
robust but slow). Poor for 1D — use `Brent` or `optimize()`.
Explore
123,283 skills indexed with the new KISS metadata standard.
robust but slow). Poor for 1D — use `Brent` or `optimize()`.
probabilities for logistic); `type = link` (default) gives on the link scale.
default 2/3); `loess` parameter is `span` (default 0.75).
y)` — cannot predict at new points.
`forward`
`SSasymp`
Type I SS (sequential) are computed — order of terms matters.
use `car::Anova()` or set contrasts to `contr.sum`/`contr.helmert`.
`quasipoisson`) allow overdispersion — no AIC is computed.
logical
test = Chisq)` for deviance-based tests; `F` is invalid for non-Gaussian families.
use the cbind trick or offset.
`factorLevelName` concatenation.
`poisson(link = log)`
but slower and with deviance-based output).
`^` means interaction crossing.
`intercept`
`contr.poly` for ordered.
`print`
gotchas
integer
not factor.
format = f
1