BookmarkSubscribeRSS Feed
Obsidian | Level 7


Before specifying spatial regression, I am at the stage of creating weight matrix of the neighbors (contiguity matrix to be precise). Now, I have two datasets:

1. Eurostat provided shapefile for NUTs level 3 that gives multiple X(long) and Y(lat) for each NUTs 3 region

2. My dataset in which I have aggregated(summed) my interest variable, A, at NUTs level 3 so that each NUTs 3 region has different no. (how many times the variable present) in it. For e.g. NUTs 3, AT112 has 15 counts of A while another NUTs 3 DE143 has 20 counts of A.

(each of these As were located at a particular location with X and Y: to that I assigned NUTs 3 by proc ginside process earlier).


When I create weight matrix from 1st dataset, it creates a list of neighbors for each X & Y pairs that are multiple for each NUTs 3. However, I'd like to have the list of neighbors for each NUTs 3 (rather than separate multiple X and Y inside each of them) to account for spatial dependence when regressing A as dependent on certain other bunch of independent variables. 


I have already referred some material on aggregation of geographical areas in spatial contiguity but that does not help me for my very specific problem. I'd appreciate any thoughts or ideas if someone has already worked on similar analysis. 

Let me know if my question is not clear. Thank you for your time.



Ready to join fellow brilliant minds for the SAS Hackathon?

Build your skills. Make connections. Enjoy creative freedom. Maybe change the world. Registration is now open through August 30th. Visit the SAS Hackathon homepage.

Register today!
What is ANOVA?

ANOVA, or Analysis Of Variance, is used to compare the averages or means of two or more populations to better understand how they differ. Watch this tutorial for more.

Find more tutorials on the SAS Users YouTube channel.

Discussion stats
  • 0 replies
  • 1 in conversation