I’m creating a map that displays US counties. While most of the counties appear correctly, I’m encountering two issues:
All counties in Connecticut are missing.
Escambia County in both Alabama and Florida is also not displaying.
I believe the geo hierarchy is set up correctly, as when I adjust the granularity, the data for Connecticut appears as expected. The datasets have also been uploaded successfully with expected of rows. However, these specific counties are missing in my output map. Could this be due to some errors in QuickSight?
Hello @Tianhe, what other levels of granularity do you have in your data besides counties? I haven’t seen anything about a specific issue with counties in Connecticut before, but I wouldn’t be surprised if there are issues with counties that have matching names if other geospatial hierarchies are missing from the data.
With some more information about the geospatial data in your dataset, I may be able to provide some more guidance on this issue. I’ll also include some documentation from AWS to map these hierarchies in your dataset.
I’ve set up a geo-hierarchy with 3 levels - country, state and county.
When clicking on ‘counties’ granularity of the hierarchy, all the counties are displayed correctly except for the cases mentioned above (e.g., Connecticut is blank).
I think hierarchy is set up correctly, because when clicking on the ‘state’ granularity of the hierarchy, I do see all the states showing up on the map, including Connecticut (which was missing when viewing at county level).
Hello @Tianhe, I am not sure that states displaying as expected would be a good indicator that the geospatial hierarchies are set up properly. Without any hierarchy set-up, QuickSight identifies US states easily, so it seems either counties aren’t properly linked to states or you may be required to add another level of granularity.
Would you be able to include something like city or ZIP code into your data as well? That should provide enough information for QuickSight to infer the missing county locations.
@DylanM Yeah I see your point. I still think it’s set up correctly though. For example, I have multiple Washington Counties (in Maine, Vermont, and Rhode Island… ), and they are all successfully displayed - so I thought the system must have had to use the linked state data to be able to show them correctly, but let me know if there is an alternative explanation…
I will try the approach for adding ZIP/city to the hierarchy (thanks!) - do I just need to add an example within the county into the new columns for QuickSight to infer the location?
Hello @Tianhe, unfortunately, I do not fully understand the intricate details of the geospatial data in QuickSight. I just know it is fairly common for people to experience issues when they are using anything more granular than Country/State that is not Latitude and Longitude.
As long as you have a single ZIP or City within each county, that should help provide the granularity required, then you can connect the hierarchies on the dataset layer. The documentation I linked above includes the details needed to establish the hierarchies. Let me know if that helps!