Spatial probit, logit, Poisson, ZIP and panel models (entry points)

Named entry points for spatial regression models that route to the estimators: SpatialProbit to the GMM spatial autoregressive probit (SpatialProbitGmm, method "gmm") or the Bayesian SAR probit sampler (SarProbitGibbs, method "bayes"); SpatialLogit to the linearized GMM spatial logit (SpatialLogitGmm); SpatialPoisson and SpatialZip to the spatial-lag Poisson and zero-inflated Poisson regressions (SarPoisson, SarZip); SpatialPanelFe to the fixed-effects maximum likelihood spatial panel models (SpatialPanelMl) and SpatialPanelRe to the random-effects spatial lag panel (SpatialPanelReLag). Identical to the Python arms morie.fn.xrspp, xrspl, xrspc, xrspz, xrspn and xrspr.

Usage

SpatialProbit(y, X, W, method = "gmm", ...)

SpatialLogit(y, X, W, Z = NULL)

SpatialZip(y, X, W, Z = NULL, rho_bounds = c(-0.99, 0.99))

SpatialPoisson(y, X, W, rho_bounds = c(-0.99, 0.99))

SpatialPanelFe(
  y,
  X,
  W,
  n_units,
  model = "lag",
  effects = "individual",
  lee_yu = FALSE,
  interval = c(-0.99, 0.99)
)

SpatialPanelRe(
  y,
  X,
  W,
  n_units,
  interval = c(-0.99, 0.99),
  tol = 1e-10,
  maxit = 500
)

Arguments

y: Outcome (0/1, counts, or a panel stacked by period).
X: Design matrix.
W: Spatial weights matrix.
method: "gmm" or "bayes" for SpatialProbit.
...: Passed to the chosen probit estimator.
Z: Instrument matrix (logit) or zero-model design (ZIP).
rho_bounds: Search interval for rho.
n_units: Number of cross-sectional units.
model: "lag", "error" or "durbin".
effects: "individual", "time" or "twoways".
lee_yu: Apply the Lee-Yu variance correction.
interval: Search interval for the spatial parameter.
tol, maxit: Convergence tolerance and iteration cap.

Value

The list returned by the underlying estimator.

References

Pinkse, J. and Slade, M. E. (1998). Contracting in space: an application of spatial statistics to discrete-choice models. Journal of Econometrics 85, 125-154.

Klier, T. and McMillen, D. P. (2008). Clustering of auto supplier plants in the United States. Journal of Business and Economic Statistics 26, 460-471.

Lambert, D. M., Brown, J. P. and Florax, R. J. G. M. (2010). A two-step estimator for a spatial lag model of counts. Regional Science and Urban Economics 40, 241-252.

Elhorst, J. P. (2003). Specification and estimation of spatial panel data models. International Regional Science Review 26, 244-268.

Examples

W <- matrix(0, 10, 10)
W[abs(row(W) - col(W)) == 1] <- 1
W <- W / rowSums(W)
X <- cbind(1, c(2, -1, 0.1, 1.5, 0.6, -0.4, 0.9, -1.3, 0.2, 1.1))
y <- c(1, 0, 1, 1, 0, 0, 1, 0, 0, 1)
SpatialProbit(y, X, W)$rho
#> [1] -0.5403822
SpatialLogit(y, X, W)$rho
#> [1] -0.5078681