0

I am using ArcGIS regression tools OLS and GWR for evaluating the association between dengue incidences and weather parameters, land use/cover, and demographic characteristics of the study area.

Poisson's regression is considered a better model for count variables as dengue cases. Can I perform the suggested analysis in ArcGIS?

  • 1
    Welcome to GIS SE. As a new user, please take the Tour. What have you tried? Boolean questions are of limited use here. Instead, explain what you have done, and what problem you've encountered. – Vince Oct 22 '17 at 13:35
  • No, in ArcGIS Poisson regression is not an option in GWR or OLS. You do need an expected Poisson distribution to your outcome for this to be a suitable model. – Jeffrey Evans Jul 9 '18 at 23:18
1

If you know python, you can convert your data to numpy arrays using RasterToNumpyArray or FeatureClassToNumPyArray.

Then, you can access the scipy module and run scipy.stats.poisson.

| improve this answer | |
0

Great question. I certainly agree with Fezter that using a programming language is a great tool in performing these types of spatial analysis. Python is fitted to work smoothly within ArcGIS assuming the IDE you are using has been synced to your ESRI account. My submissions (seen below) uses the RStudio IDE to create a multivariate Poisson regression analysis for broadband internet acquisition in Kentucky counties with x variables as percentiles of college graduates, age, and unemployment rates.

As it pertains to running a regression analysis in R, the 'GIS Tools' package can create the linear models which produce the descriptive statistics often cited in a Poisson analysis.

library(GISTools)
library(sf)

#connect the broadband info from the excel sheet into an sf object spatial #dataframe 'KY_df'

KY_df <- inner_join(Ky_counties, KY_bb, by=c("NAMELSAD10"="ck_name"))

#fit a linear model, where the first argument is your dependent variables #followed by your explanatory variables

mod1 <- lm(adopt05~college + unemp, data = KY_df)

#view your results!
summary(mod1)

Again, using R or any programming language can be difficult at first but with every new exercise you get more experience and better results. I hope this is helpful!

Happy coding

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.