sarpoisson: Spatial Autoregressive Poisson Model
In gregmacfarlane/sppois: Spatial Autoregressive Poisson Modeling

Description Usage Arguments Value See Also Examples

This function implements a limited-information maximum likelihood estimator for Poisson regression models. The estimator was described by Lambert, Brown, and Florax (2010).

1	sarpoisson(formula, data = list(), listw = NULL, method = "liml", ...)

`formula`	A symbolic description of the model to be fit. The details of model specification are given for `lm`
`data`	An optional data frame containing the variables in the model. By default the variables are taken from the environment which the function is called.
`listw`	A listw object created for example by `nb2listw`
`method`	The method to be used for fitting the regression equation. Defaults to "liml", a limited-information maximum likelihood. Other options are "fiml" (full-information maximum likelihood), "model.matrix" to return a model matrix, and "non-spatial", which will execute a non-spatial Poisson regression (identical to `glm`).
`...`	Further arguments passed to nlm.

A list of class 'sarpoisson' containing the following components:

coefficients: The estimated coefficient values.
fitted.values: The estimated mean of the poisson distribution.
residuals: Difference between estimated mean and observed value.
df.residuals: Degrees of freedom remaining in residuals.
df.null: Degrees of freedom in the null model.
logLik: Numerical log likelihood.
rank: Rank of the model.
call: Model estimation call.
nlm_results: Complete results from nlm.
information.matrix: Fischer information matrix, obtained as inverse of Hessian.
method: The method used in maximum likelihood estimation.
xlevels
terms: Terms of the model frame

lagsarlm

  summary(sarpoisson(crime_i ~ income + home_value, data = columbus_crime,
          method = "non-spatial"))
  summary(sarpoisson(crime_i ~ income + home_value, data = columbus_crime,
          listw = columbus_neighbors, method = "fiml"))