This is possible with datablending 2 sources but unfortunately I have not tried this with 3 datasources.
But what you will be delighted about is the new feature in Tableau 10 that allows cross data joins
allowing you to link your 2 databases and excel in one single datasource.
check out the link above for more info.
Many thanks for your speedy reply. That's great news about Tableau 10. Unfortunately I don't think we will be upgrading to Tableau 10 for a while.
Please would you be able to explain to me how to achieve the merge with 2 data sources and we will go with that for the time being.
[Field Name from Primary] + [Secondarydatasource].[Field from secondarydatasource]
this is all that is required as far as you have defined your relationships between the datasources
In the calculated field window when you start typing fieldname it will give you options of field names in secondary datasource you can just select the field.
Thanks for the suggested solution unfortunately - I'm getting the following error when trying to achieve the blend.
All fields be aggregate or constant when using table calculation functions or fields from multiple data sources.
Please can you advise on this error?
try using ATTR([field name))
Hi Khalid - I've tried ATTR fix as you suggested - but unfortunately just get a NULL value returned.
Do you have any other suggestions - Basically just got a list of names in one data source and list of names in the other and want to make a big single list of data from both sources?
Seems to be a problem as the data is string?
How is the relationship between the 2 data sources defined
Is it possible to attach a sample workbook
You might try "bridging" the data using a relational database.
I create dynamic links to my different data types in a single relational database. As an example, I have Excel files, ODBC tables, and Access tables all linked to an Access database. (I'm not sure, but I think you can do this with an MS Sql Table, as well).
When I use Tableau to link to the database, I can join all these seemingly disparate types on the primary data screen. I call this "bridging" but I'm sure there's a more accurate name.
I won't say this is a silver bullet-- there may be performance issues this causes (though creating an extract may mitigate them).
It may also exclude or limit the information you need-- example: if each of your files had a different server name, and you're seeking to get the union of all of them, I'm not sure how easy that would be.
Hopefully this will give you another way of looking at things.