Hi All,   Sorry for the delay in updating this post, I've been down in Fiordland, one of our fantastic national parks, blissfully out of touch.   Thank you very much for your help and ideas. As you will see, I have flagged Fredrik's second suggestion, using Display Rules to colour specific columns, as the solution to my original question. Ideally though, I would like the Crosstab object in Visual Analytics to behave a bit more like Proc TABULATE in terms of nesting of columns. For the moment. though, I have a solution that I can put forward to my colleagues and hopefully make them happy or at least happier.   Once I've unpacked my rucsac and cleaned my boots I aim to put an entry on the SAS Ballot page of Communities to try to lobby support to get the Crosstab object in SAS Visual Analytics improved somewhat.   Here's a picture of one of Fiordland's cheeky mountain parrots, the Kea.   Cheers,   Downunder Dave. Wellington ... View more

Thanks for following up, Fredrik. This sort of error is a bit of a blind spot for many of us. ... View more

And another one   data have; x='PraveenKumar';output; x='RamSita';output; x='JorgePaul';output; run; data want; set have; First_name=substr(x,anyupper (x),anyupper (x,2)-1); Last_name=substr(x,anyupper(x,2)); run; ... View more

Thank you so much Mr.Fredrike.. Finally, 'case when' statement is very helpful. I just forgot that the record is case sensitive. The real record is 'Quarterly' not 'quarterly' so that I got the error. So, I change the first alphabets to upper case, then it works 😄         ... View more

@IgawaKei29 wrote:   The dataset is roughly 1mil rows so that adds to my complexity with trying to work this out.    Don't use the original dataset for testing. Just create a temporary dataset consisting of a few representative observations from it and develop the code using this small dataset.   Just to add to @ballardw's suggestion: If your replacement strings are longer than the original ones (as is the case in two of your examples), you need to make sure that the lengths of the respective character variables are sufficient to accommodate the longer values. If they are not, redefine the lengths as shown in the example below to avoid truncation: data have; color='Blue'; run; data want; length color $5; /* important: LENGTH statement before SET statement */ set have; if color='Blue' then color='Green'; run; ... View more

Hi @upadhi,   To add to @Joy_SAS and @FredrikE's responses, for future reference there is a series of articles, called D3 Thursday, about the Data-Driven Content object in VA 8.2.   Anna ... View more

Ah yes, that works. Didn't think to do the counts in VA. Perfect!   Just for reference, here is the actual text that is working for me:   IF ( 'STATUS'n In ('SOLD') ) RETURN 'COUNT'n ELSE 0 ... View more