Functions for binary choice example in the vignette.
binary.f(P, data, priors, order.row = FALSE)
binary.grad(P, data, priors, order.row = FALSE)
binary.hess(P, data, priors, order.row = FALSE)
binary.sim(N, k, T)| P | Numeric vector of length \((N+1)k\). First \(Nk\) elements are heterogeneous coefficients. The remaining k elements are population parameters. |
|---|---|
| data | Named list of data matrices Y and X, and choice count integer T |
| priors | Named list of matrices inv.Omega and inv.A. |
| order.row | Determines order of heterogeneous coefficients in parameter vector. If TRUE, heterogeneous coefficients are ordered by unit. If FALSE, they are ordered by covariate. |
| N | Number of heterogeneous units |
| k | Number of heterogeneous parameters |
| T | Observations per household |
For binary.f, binary.df and binary.hess, the log posterior density, gradient and Hessian, respectively. The Hessian is a dgCMatrix object. binary.sim returns a list with simulated Y and X, and the input T.
These functions are used by the heterogeneous binary choice example in the vignette. There are N heterogeneous units, each making T binary choices. The choice probabilities depend on k covariates. binary.sim simulates a dataset suitable for running the example.