With ArcGIS, it looks that we can directly use ESRI GWR model (rather than R GWmodel) to effectively do spatial regression for some applications ...
With this GP tool, how can we set up those non-Euclidean parameters in GP tool (ESRI GWR)?
For example, in the many cases like house price prediction in real estate market, the non-Euclidean variables (like network distance and travel time in house market) should be considered for more reliable analysis and prediction ……. (another example is, ESP failure analysis in O&G)
Secondly, how can we integrate a variable like house view (House_View) into ESRI GWR?
As well known, the price of the houses with better house view (saying, a house in the hills facing the remote mountain, lake, downtown …) actually will have much better potential …
What about the combination of the raster like the house view raster, average annual temperature raster, or average rainfalls raster with ESRI GWR workflow?