Claude, is the issue that the various geo roles are all in the same column in the data (that is, FR and CH011 are both in one column)?
If that is the case, you need a way to differentiate the roles so that Tableau can consume them. Fixing it in the source data is always best, but often not possible.
You (probably) can create something with a set of calculated fields inside Tableau, or possibly with some custom SQL depending on the data source.
Thanks for the reply, Michael!
Yes, the different locations are in the same column. entiteled "location". This wold not be a problem to spred this into more columns, but would it work? in some "country code", I will not have a detailed location and the column would contain a "null" t this place.
Il my simple tests, the date are not displayed.
What I would like to avoid is to introduce every info since, taken separatly in Tableau, they can be recognized... But not together :-(
It's a tricky problem, I agree. I'd have to play around with the data to be certain of how to handle it, but it would be simpler if the following is true.
For a given country code and "domain", do you always have the same geodata granularity (e.g., for CH, do you have Canton level data for the whole data set?), or is it sometimes yes, sometimes no?
By "domain", I mean population data vs. income data, for example. It might be OK if you have one level of granularity for population but a different one for income, or it might not (that depends on what you want to visualize).
The only way you're going to get Tableau to be able to assign a geo role to the data is if you have a distinct column for each geo role type in the data, somehow. You COULD, possibly, create a column that explicitly states what the geo role is for a given row and use that to create a new calculated field that gets you what you need...
For each "location" I have one value: the number of items at this place. In fact those are for the location of people's job at a moment. 5 are in France, 2 are in US-NY, 6 are in CH040, etc etc. I simply imagined 2 columns: 1 "location" ; 2 "number".. but this does not work . Because they are not at the same level... If I split in several columns, tins does not help, as mentioned earlier.
Ah, I see! Yes, that does make it difficult. Very hard to do without a consistent level of detail.
What about latitude/longitude values for location instead? Are all the locations at consistent physical locations, or are these individual residences or hotels or other temporary places that can change?
Are you literally trying to show a map of the world with pins for each employee?
I would like precisely to avoid the introduction of each coordinates since all those regions/states are available in Tableau (for some of them, polygons, for some it looks like the center of gravity of the country). The idea is to display the number of people (not pins) by country or in more detail by region when available (USA and CH for the moment, France & Germany later).
2 of 2 people found this helpful
Tableau only supports geocoding one field at a time and a field can only have one geographic role. So if you are looking to make a map with multiple levels, you can not easily do this today. My recommendation would be to compile a geography dataset from GADM, EuroStat and US Census using a GIS tool like QGIS or a data preparation tool like Alteryx or FME. The output from this would be a single spatial file with all of the relevant geographies.
Thanks a lot for your precise answer. I will try your recommendation and hope that in future versions Tableau will support multiple resolutions (and norms) of geo localizations. In a global World, this seems to be an idea.
Thanks again and best wishes,
Has there been any progress done from the mapping team so that we are able to compute the coordinates for multiple geographical roles at the same time, as per Claude's original post?
1 of 1 people found this helpful
No, that would still be a function for a GIS application. Tableau cannot parse different geo roles that exist in the same column of the data.
If you have multiple columns with different geo roles, then it works fine.