Hi there,
I'm using the hot spot analysis to assess clustering/dispersion of of hot and cold fire spots and have generated a spatial weights matrix (.swm), however my weight column indicates the distance between points as 1 for all the point features. Please note I am using a projected data.
Any comments as to why, and how to trouble shoot this issue would be most appreciated.
Thanks in advance.
Solved! Go to Solution.
Well...from the help file
At the most basic level, there are two strategies for creating weights to quantify the relationships among data features: binary or variable weighting. For binary strategies (fixed distance, K nearest neighbors, Delaunay Triangulation, contiguity, or space-time window) a feature is either a neighbor (1) or it is not (0).
From here
Generate Spatial Weights Matrix—Help | ArcGIS for Desktop what options did you use.
How Generate Spatial Weights Matrix works—Help | ArcGIS for Desktop
or were you using something else?
Do you want this directed to the Spatial Statistics or Spatial Analysis place?
Yes, I'm using the links you provided. Please could you direct this question to the spatial statistics/spatial analysis group.
Thanks.
What were the parameters used? This has a large impact on results.
And you are absolutely sure that you have a projected file.
An image would help as well of the area and the table
Yes, projected data. Please see image of spreadsheet below and output of (.swm).
Parameters used: Fixed_Distance /Euclidean
Threshold distance of 4500 m (decision based on the output of ''calculate distance band from neighbor'' (based on 8 neighbours).
|
Field1 | UNIQUE1_ID | NID | WEIGHT |
0 | 7945 | 818 | 1 |
0 | 7945 | 13443 | 1 |
0 | 7945 | 11107 | 1 |
0 | 7945 | 3721 | 1 |
0 | 7945 | 7961 | 1 |
0 | 7945 | 11104 | 1 |
0 | 7945 | 11108 | 1 |
0 | 7945 | 7984 | 1 |
0 | 7961 | 11107 | 1 |
0 | 7961 | 13443 | 1 |
0 | 7961 | 5875 | 1 |
0 | 7961 | 3721 | 1 |
0 | 7961 | 833 | 1 |
Thanks Dan.
Well...from the help file
At the most basic level, there are two strategies for creating weights to quantify the relationships among data features: binary or variable weighting. For binary strategies (fixed distance, K nearest neighbors, Delaunay Triangulation, contiguity, or space-time window) a feature is either a neighbor (1) or it is not (0).
Many thanks Dan. I was concerned as all the feature points have neighbours in that case - as there are no 0. Will double check.
You might want to check your threshold distance then or try another model. Post a map of your distribution if you want.
Glynis, could you mark this question by checking the answer that worked, or assumed answered or convert to a discussion. Threads with closure indicate the status and are useful to others when looking for help on similar topics.
Hi Dan - thanks. Here is the distribution map - succeeded in generating a .swm - I verified my data based on a re-run of my data in GeoDa software and the same patterns emerge. Thanks for you help.