fmess_ar2.m

空间统计工具箱

function [bmax, srds, prhigher, emax, maxlik]=fmess_ar2(x, y, d)
%
% [bmax, srds, prhigher, emax, maxlik]=fmess_ar2(x, y, d)
%
%This function estimates an autoregressive matrix exponential spatial
%specification (MESS).
%
%
%INPUT:
%
%x contains n observations on k independent variables (n by k)
%
%y contains n observations on a single dependent variable (n by 1)
%
%d is a n by n spatial weight matrix (n by n)
%
%
%OUTPUT:
%
%bmax is the k+1 by 1 vector of parameter estimates with the SAR regression parameter estimates as the first k elements and 
%alphamax, the global spatial dependence parameter (a scalar), as the last element.
%
%srds is the k+1 by 1 vector of signed root deviances associated with bmax. These have a t-ratio like interpretation. 
%The first k elements are for the delete-1 hypotheses and the last is for alpha=0. Squaring the srds yields the likelihood ratios.
%
%prhighers is a k+1 by 1 vector of estimates of obtaining a higher likelihood ratio under repeated sampling.
%
%emax is n by 1 column vector of MESS residuals.
%
%maxlik is a scalar giving the maximum of the profile log-likelihood
%
%
%NOTES:
%
%If you have extremely large spatial dependence parameters such as alphamax
%of over 6, you may need to increase nq.
%
%If you use this function, please cite: 
%
%You can see the following for more information:
%
%LeSage, James and R. Kelley Pace, 揝patial Dependence in Data Mining,