A model for odds linear in some feature.
as.linodds(object, formula, beta)# S3 method for linodds
predict(object, newdata, type = c("eta", "mu",
"erank"), na.action = na.pass, group = NULL, ...)
# S3 method for linodds
coef(object, ...)
some list-like object.
an object of class "formula" (or one that
can be coerced to that class): a symbolic description of the
model to be fitted. The details of model specification are given
under ‘Details’.
the fit coefficients.
a data.frame from which we can extract a model
frame via the formula of the object.
indicates which prediction should be returned:
etaThe odds.
muThe probability.
erankThe expected rank.
How to deal with missing values in y, g,
X, wt, eta0.
the string name of the group variable in the data, or a bare character with the group name. The group indices need not be integers, but that is more efficient. They need not be sorted.
other arguments.
An object which holds a formula, some fit coefficients
\(\beta\) which fit in that formula to generate odds
in odds space.
The odds can then be converted, via predict.linodds
to probabilities,
or to expected ranks under the Harville model.
Both harsm and hensm return
objects of class linodds.
We think of linear odds as \(\eta = x^{\top}\beta\), for independent variables \(x\). The odds, \(\eta\) are converted to probabilities, \(\mu\) via \(\mu = c \exp{\eta},\) where the constant \(c\) is chosen so the \(\mu\) for a given matching sum to one.