Hi Dom,

If your data sources are relational ones,

you'd be better to create a CatalogName field

in each of your data sources via Custom SQL.

There are two ways of doing it:

1) From the Datasource UI choose:

Data --> Convert to Custom SQL

Then edit the first line of your SQL statement like this:

SELECT 'BITS' as CustomerName,

<your code follows>

2) Re-create your data source schema from scratch.

When placing the first (fact) table / view on a pane, choose:

Data -->  Convert to Custom SQL and edit your code as above.

Then join other tables / views / Custom SQL selects.

Hope this makes sense.

Yours,

Yuri

Data -->  Convert to Custom SQL and edit your code as above.

Nice tip. Strangely, I cannot remember seeing it before. I guess it is either a relatively new feature or forgotten because it is so long time since I used Tableau. This is the only official writing about it which I have been able to find until now:  Converting a Connection to Custom SQL | Tableau Software

That said, there are several other answers/articles that mention Convert to Custom SQL:

Re: Hello, in version 8.2, where is the custom sql option in the excel workbook connection ?

Re: 8.2 New Connection has Window Less Functionality?

Self Blends and Data Padding - The Information Lab

Tableau Doctor: 1 Column, 2 Dates? Use Custom SQL | Tableau Software

https://www.google.com/search?q=%22Convert+to+Custom+SQL%22+site%3Atableau.com

Yuriy,

This is an interesting feature I didn't know.

Still this won't help him because this doesn't allow union from multiple data sources.

Scaffolding is a way to help him to some extent.

Vizible Difference: Blending Data Via Multi-Dimensional Scaffolding

Hi, Alexander,

If I understood correctly, Dom has been trying

to append an extract based on a "first DS" (BITS_Extract.tde)

with data from an extract based on a "second DS" (CS_Extract.tde) --

the same as described in Tableau Desktop Online Help:

http://onlinehelp.tableau.com/current/pro/online/mac/en-us/help.htm#extracting_addfromfile.html

Doing just that is not a UNION per se,

it is simply adding a next portion of data,

so the CatalogName field should be already

materialized in each extract before appending.

The latter is not possible

with a row-level single-value "calculation".

Correct me if I'm wrong.

Yours,

Yuri

I see. Thanks!

This will enable union across multiple data sources, it seems.

Thinking of it a little bit more,

there are cases when it would be better

to preserve some properties of datasources

when doing "light" transformations in Tableau --

such as having Row_ID for a Pivot,

or Datasource_ID for an upcoming 9.2 Union.

Just my 0.02$

Yuri,

It seems that such appended data is a one-time thing. It can't be made dynamic. That means, if I added data from Source A to Extract AAA, then refreshing AAA will not add data from Source A again.

So the append is materialized. It is not a union on the fly.

If the extract AAA is an extract from a database, after appending data from a file or a data source, it can keep adding data by incremental refresh from the original database.

Yep, append from file/extract

follows the same logic

as an incremental refresh.

Doing full refresh removes

all appends / increments.

Yuri,

I found that Convert To SQL doesn't apply to extracts from Excel files.

Do you know why?

Maybe that there is already an option to do that via legacy connection.

Alexander,

It is my understanding that Tableau exposes Custom SQL interface

for each database engine that is able to respond to SQL queries -- 

be it ANSI SQL or the likes (bQL for Google BigQuery or HiveQL for Hive).

The Legacy Connector for Text / Excel uses plain old

MS Jet Database Engine, which talks plain SQL, too.

That's why one could use a Custom SQL interface,

hence Convert to Custom SQL.

"Native" Text / Excel connector -- new as for the version 8.2,

and in particular with the introduction of Tableau for Mac -- talks no SQL.

Hence no Custom SQL / Covert to Custom SQL.

(And no non-equi-joins :-)

I am thinking of "native" connector as an advanced (and portable)

parser / loader for Tableau Data Engine. I am speculating further

that common core is shared among other connectors as well --

R/SAS/SPSS - Google Analytics - Web Data Connector.

Yours,

Yuri

Yuri,

This gentleman can apply "convert to custom sql" to excel file.

Self Blends and Data Padding - The Information Lab

I can't. I am using 9.1. and the option is grayed when the file is excel.

See if you can reproduce it?

His screenshot:

My screenshot:

Hi Alexander,

Citing this gentleman's blog post (one paragraph above the picture):

For this example with Superstore Sales, the mention of Custom SQL comes with a caveat, and that is that I’m going to have to use the Legacy Excel connector to take advantage of JET, and as a result I’m afraid this is going to be PC only. If you’re on MAC and have data in Excel you’re going to have to do some pre-processing to either get it into a database or utilise a product like Alteryx.

To recap the legacy connecter, the trick is to select the appropriate file “open” option when you locate your Excel file:

So the pre-requisites are:

-- Windows

-- Tableau Desktop

-- MS Jet Database Driver (all versions or all versions)

-- When opening Excel / CSV file, choose Open with Legacy Connection

Yours,

Yuri

Yuri, I did see that passage.

But it seems his screenshot shows he is using Data>Convert to Custom SQL to

implement his SQL code.

Could you reproduce that with an excel file?

On Sun, Nov 1, 2015 at 11:46 PM, Yuriy Fal <