I implemented the hotspot analysis using ArcGIS which relies on the Getis-Ord Gi* statistics. Based on the information reported on the corresponding ESRI webpage, even if data are clustered the presence of one feature showing a high or low z-score value do not indicates the presence of a significant hot or coldspot. In fact, to be considered as a statistically significant spot, a feature must have a high/low value and be surrounded by other features presenting high/low values as well.
However, what is the exact number of features required to define the presence of an hot or coldspot? two, three, four, etc.? or simply more than one?
Thanks in advance.
Dear Marcello,
1) take a look at section 3.9 of the following book
Statistical Detection and Surveillance of Geographic Clusters
By Peter Rogerson, Ikuho Yamada
2) you may consult also R packages to calculate Getis and Ord's statistic and test its significance among them "rgeoda".
https://github.com/geodacenter/rgeoda/
and
https://cran.r-project.org/web/views/Spatial.html
(search Getis-Ord G inside pages of the above link)
Sorry, I am not enough well practiced with ArcGIS.
Yours,
Hamid
WELCOME Marcello,
IF POSSIBLE PLEASE leave a message for me if you find your answer. I am interested. Thanks
- Badges
- Science topic
- Similar topics
- Correction