fdet_interp_seq2.m

空间统计工具箱

function [detvalz, alphafine]=fdet_interp_seq2(d, sym, alphacoarse, alphafine)
%
%[detvalz, alphafine]=fdet_interp_seq2(d, sym, alphacoarse, alphafine)
%
%This function computes ln|I-a*d(1:i,1:i)| for i=1 to n for a grid of values of a (stored in alphafine) where d is an
%n by n spatial weight matrix.
%
%
%INPUT:
%
%The n by n spatial weight matrix d is a required input.
%
%The optional scalar sym gives whether a matrix is asymmetric (sym=0) or symmetric (sym=1). Not supplying this forces
%the routine to test for symmetry which could cost time for large problems.
%
%The optional itersub by 1 vector alphacoarse gives the actual evaluation points of the log-determinant for the 
%corresponding values of a in alphacoarse. If you supply alphacoarse, you must supply alphafine.
%
%The optional aiter by 1 vector alphafine gives the arguments associated with the output log-determinant values. 
%If you supply alphafine, you must supply alphacoarse.
%
%
%OUTPUT:
%
%detvalz is a iter by n matrix where the first column is a vector of the
%values of the spatial dependence parameter (equal to alphafine when input), and the second column is a
%vector of the log-determinants.
%
%The aiter by 1 vector alphafine gives the arguments associated with the output log-determinant values. 
%
%
%NOTES:
%
%This function was discussed in:
%
%Pace, R. Kelley, and James LeSage, 揝patial Autoregressive Local Estimation,