Extract coefficients and other estimates from mnprobit object.
# S3 method for mnprobit
coef(object, ..., alt = NULL, regime = NULL, type = "coef")See 'Details' section.
object of class "mnprobit"
further arguments (currently ignored)
integer representing index of the alternative
integer representing regime of the continuous equation
character representing the type of the output. Possible options
are "coef", "coef2", "cov1", "var",
"cov2", coef_lambda.
See 'Details' for additional information.
Consider notations from the 'Details' section of
mnprobit.
Suppose that type = "coef". Then estimates of \(\gamma_{j}\)
coefficients are returned for each \(j\in\{1,...,J\}\). If alt = j
then only estimates of \(\gamma_{j}\) coefficients are returned.
Suppose that type = "coef2". Then estimates of \(\beta_{r}\)
coefficients are returned for each \(r\in\{0,...,R - 1\}\).
If regime = r then estimates only for the \(r\)-th regime are
returned.
Suppose that type = "cov1". Then estimate of the covariance matrix of
\(u_{i}\) is returned. If alt = c(a, b) then the function returns
\((a, b)\)-th element of this matrix i.e. an element from
a-th row and b-th column that is an estimate of
\(Cov(u_{ai},u_{bi})\).
Suppose that type = "cov12". Then estimates of covariances between
\(u_{i}\) and \(\varepsilon_{i}\) are returned.
If alt = j and regime = r then the function returns
an estimate of \(Cov(u_{ji}, \varepsilon_{ri})\).
Suppose that type = "var" or type = "cov2". Then estimates of
the variances of \(\varepsilon_{i}\) are returned.
If regime = r then estimate of \(Var(\varepsilon_{ri})\)
is returned.
Suppose that type = "coef_lambda". Then estimates of the coefficients
for \(\hat{\lambda}^{t}_{ji}\) are returned i.e.
estimates of \(\tau_{jt}\) for each regime.
If regime = r then estimates are returned for the \(r\)-th regime.
If in addition alt = j then only estimates for this \(j\)-th
alternative and \(r\)-th regime are returned.