I am sorry if this is a simple question. I am trying to conduct a GWR for hedonic analysis of hotel room rates. For OLS I have conducted analysis including different room types (e.g. double, suite, twin, etc.), but while doing GWR with the same data I get error on ArcGIS that in some places there are more than 80 coincident data values. This may be due to the fact that a single hotel location may contain several superimposed point features (different room rates according to room types).
I'm not entirely sure how to fix this problem. Should I just limit my GWR consideration to particular types of rooms and reduce the sample size, or should I use a different method?
Any help would be appreciated.
Are you convert feature layer to the raster layer and you should reclassify
This problem is rather complicated. The problem is from programming error. The programmer did not take all the scenario into consideration. That is why OLS works when the computer application does not. Any misbehaviour like this should be reported to the software producers.
Clearly the coincident cases are not actually coincident (unless a tower block) but the spatial resolution of the data is too coarse to distinguish differences. You could therefore 'jitter' the geographical coordinates with say random uniform numbers to the next two decimal values , or a number small than the difference to the next nearest place . R has a function for this https://stat.ethz.ch/R-manual/R-devel/library/base/html/jitter.html . Or you could input one summary at that point, say a median.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size