5 Replies Latest reply on Mar 28, 2016 11:02 AM by Thomas Exter

    Combining "Mark" Maps

    Luke B



      I am trying to create a map that plots two different types of marks, where each mark has its own lat/long, but the two different "types" will be displayed according to an attributed value. This is a poor issue statement, but the following example provides better clarity.


      The first mark will be for a county population, where the mark size is dependent on the county's population. The second mark, a gauge location, will be shaded based on the associated value from a column associated with each gauge. The intention of the visualization is to highlight the relationship between the color shading of the gauge mark, relative to the nearby population. I may be interested in using shaded polygons instead of a single point for each county.


      The attached workbook contains the desired outcome I am seeking for each of these visualizations. I am simply unable to combine the visualizations into a single map.


      I have read numerous threads about the importance of combining data sets, but am struggling to adapt the solution to my problem.


      Map 1:



      Map 2:



      Thanks in advance

        • 1. Re: Combining "Mark" Maps
          Mosin K

          I am sure the gurus here will have better ways of doing this, all I could think of:

          - Use your county data to extract the shape files from US Census county data.

          - Append population info to the shape files

          - Merge the Gauge info with the county info.

          - Plot as Dual-Axis map plot.


          Hope this helps.


          1 of 1 people found this helpful
          • 2. Re: Combining "Mark" Maps
            Mahfooj Khan



            You're trying to display measures from two different sources.

            You have two options one either merge your both the data at source level like plotting gauge in population data or you can use data blending to join both the sources. But in data blending you should have blending key to join both the sources. Here I didn't find any common filed so that you can blend the data.

            Once you come up with how to link both, use dual axis map to display your measures.



            • 3. Re: Combining "Mark" Maps
              Mosin K

              I dont think we need a SQL like Join here. A Union will do just fine - only make sure that we align the columns by Name and that Latitude/Longitude line and that there is a Classifier for each row which indicates if Lat/Long belong to County or Gauge.


              That said, the Area names here are County names. Changing the Area to 'Geographical Type County' while generate the County Filled Map, it does not generate it for all the counties.


              It appears that tableau is unable to decode the county names as unique. To overcome that, I created a new State Variable and labeled is "CO" (all counties appear to be in Colorado) and placing the "Area" measure under State hierarchy. This gives much better results and I think is simpler than custom shape job I had done earlier.


              The dashboard below compares the results of dual axis plot with;

              - Custom Extracted Shapefile (no missing counties but a large dataset)

              - Just Converting Area to "County" geography (missing counties)

              - Creating State to County hierarchy (no missing counties and smaller dataset)



              1 of 1 people found this helpful
              • 4. Re: Combining "Mark" Maps
                Luke B

                Thank you so much for your work on this. This appears to have the functionality I am seeking.

                • 5. Re: Combining "Mark" Maps
                  Thomas Exter

                  Hi all, just to point out that the well-known state-county issue can also be handled by editing the "unknowns" then "edit locations" then specify state = Colorado. Or, adding both state and county to the detail shelf so that the built-in hierarchy is utilized. Still, your workaround is interesting.


                  Best regards,