Re: [Maptitude] calculating spatial lags sums...

Fri, 16 Jun 2006 16:53:06 -0700

Hi Dave,
 
Thanks for the fast prompt reply. I had tried the adjacency route and it works. However, Maptitude does not generate adjacency matrices (unless one uses GISDK). In Maptitude I tried the following and I am unsure if my results will be correct. It is:
 
Create a band on the 304 BG layer (parent layer) with a 0.01mile buffer and check the 'separate band' box and save it as the band layer (child layer). Now in the new band (or child) layer create a field and  (rightclick) 'Edit-->Fill-->aggregate' the column with the original BG (or parent) layer field that contains the # of bars. Since the band (or child) layer is slightly larger than the BG (or parent) layer it picks up the values of the 'adjacent' units in the parent layer. Spot checking worked but I am not sure if it is robust enough.
 
Regards, Rudy B.
 


 
On 6/16/06, Dave Twehues <[EMAIL PROTECTED]> wrote:

Rudy,
One idea (maybe not simplest) would be to use the spatial adjacency matrix as it will return binary if touching or not and just multiply by alcohol units through the matrix and sum...then for n-order 'lag' effects you would take the first matrix and if value is 1 do the matirx on that (so there will be as many n-order matrices as number of neighbors for lag n-1) removing duplicates as you go as there should be a lot...etc...this is very generalized description obviously but, if you go back through the postings there was a thread about creating the Queens and Rook spatial adjacency matrix and that should get you started...1st order is easy...just code a loop and duplcate removval for n-order (complexity will increase fast but with only 304 it should be negligable)....hope this gets you jump started towards a faster solution...Dave

From: [email protected] [mailto: [email protected]] On Behalf Of Aniruddha Banerjee
Sent: Friday, June 16, 2006 6:39 PM
To: [email protected]
Subject: [Maptitude] calculating spatial lags sums...

 

Hi,

I have 304 block groups with # of alcohol outlets as a variable. Now I
want to calculate another variable that is a sum of itself and all its
immediate neighbors. Let's call it the first lag variable. Also is it
possible to calculate 2nd 3rd 4th etc lagged effects? What's the
fastest way to do this? Regards, Rudy B.


__._,_.___


SPONSORED LINKS
Geographic information systems Geography game Geography

YAHOO! GROUPS LINKS



__,_._,___

Reply via email to