is the source for this data the same data source? This example is a Excel worksheet which I could work with to provide a potential solution but if the data sources are different this would not be applicable or useful.
One other thought is to utilize custom geocoding for the [System] locations.
After playing with this a bit more, I managed to create a view using a Dual Axis. check out the Sheet "Dual Axis" in the attached. I created a field for Hospitals in the primary data source from the second date source and linked the two by zip code. I then took all the null/* zip code values and made teh color for thes a grey color. I know Drummey has done this in the past by making the color transparent which would work even better. I am linking Jonathan Drummey in since I can not find that post on Drawing with Numbers but I recall him doing this in the past.
Map sample_PV.twbx 243.3 KB
1 of 1 people found this helpful
I'm not remembering any DwN post on this, there have been some different forum threads. Given the two data sources and using Tableau's built-in geocoding then the only option is a data blend. To hide the extraneous marks I use a transparent 1x1 pixel PNG for the ones I want hidden:
However, there is one problem with the data that might require the data to be prepared in an alternative way - a number of zip codes have multiple hospitals:
Because data blending occurs *after* aggregation, the Calculation1 for showing hospitals is turning up * for those zip codes that have multiple hospitals:
The only workaround to get a view with all zip codes from the Share % data and all hospitals would be to do a union of the two sources. We can use the legacy connector for Excel sources and custom SQL to do a union or do the union outside of Tableau. (version 9.3 that is currently in beta supports unions for Excel & text files).
An advantage to this is that we then don't need the transparent PNG shape mark hack and instead can build a dual axis map where we control the level of detail on each axis so Tableau is only drawing as many marks as we actually want to see (though it does require extra work to make sure the measure results come out right).
That still doesn't quite solve the problem of displaying all those hospitals that are in the same zip code. My suggestion for that (given the union) would be to use a pie chart, since given the density of points the extra work to jitter them would be quite difficult.
Can this be saved as a file that can be used in Public? We are on 9.1 here, so I can't open it. Really wish Tableau would fix that.