nb_mat {starma}

Neighbourhood weight matrices for France's 94 departments

Description

This data file provides three neighbourhoods for the 94 metropolitan French departments:

- dlist: distance-based neighbourhoods; two departments are considered neighbours if their centroids are within range of 100km.

- klist: four closest neighbours; each department is connected to its four closest neighbours, the distance being calculated between centroids.

- blist: common border neighbours; two departments are considered neighbours if they share a border.

These neighbourhoods are designed to be used within the starma-package. First element is the identity matrix (0-th order neighbours). Second element is the common border contingency matrix of the department (1-st order neighbours). Elements three to five are the weight matrices lagged from the previous one (2-nd to 4-th order neighbours).

They have been computed used the package spdep and its functions readShapePoly, poly2nb and nblag.

Usage

nb_mat

Format

Three lists of 5 weight matrices, of dimension 94x94