Thanks for reply, Alex!
Actually, my data are too big to fit in excel.
And they are long and narrow, its a big table of more than 1M customers (rows) and 75 columns.
Any ideas? I can't believe that such a powerful BI tool like Tableau cant calculate the maximum value in a row.
Transposition of the base is not an option because it consists of hundreds of thousands users, and we need to aggregate on each of them.
You could try custom SQL - the second answer looks like it may work for you if you're on SQL of some sort: http://stackoverflow.com/questions/71022/sql-max-of-multiple-columns
If you're coming from a text file I don't know if that will work.
You may be able to use RAWSQL as a calculated field as well, depends on your source: http://kb.tableausoftware.com/articles/knowledgebase/raw-sql
You can also use nested if statements, but I doubt that will work with the number of columns you have.
You wrote, "I can't believe that such a powerful BI tool like Tableau can't calculate the maximum value in a row." A helpful frame of reference for me as I've had some of the same reactions is the following: Tableau fits the mold of a Clay Christensen-style "disruptive" entrant into the BI market, in the sense that it does some things way better than vendors, a number of things not as well as other vendors, and totally doesn't do some things other vendors do. As far as I'm concerned, Tableau is the best by far at interactive data visualization and ease/speed of use, and wins on that, and part of the cost is that in cases like yours we have to resort to various workarounds to get the calculations we want.
In this case, Tableau chose to focus on working better with tall vs. wide data, and that instead of building out every conceivable calculation themselves, they would rely on the underlying data source for as much of the calculations as possible.
Alex, thanks again! I have the CSV file as the source.
But still, even for SQL, thats OK if you have some 3-5 columns, but if you have 75 of them?? It'll be tricky.
As an addendum, in the responses I've seen about trying to do this sort of calculation, most point that if you're even trying to do this, there are likely issues in the database design - i.e. it is not appropriately normalized - in other words you are trying to do something that is out of the bounds of normal use.
You may want to look at PowerPivot - I don't know if that will let you look across a wide file with that many rows: http://www.microsoft.com/en-us/bi/powerpivot.aspx
You could also look at writing your own code in Python or similar to add a max column
I'll try to use PowerPivot, hope it'll help!
As for db normalization thats ok if you dealing with the relational structure. But I dealing just with the plain raw data on transaction frequencies.
Jonathan, I personally like Tableau a lot, just want to add some flexibility.