Hi, Just wondering if it is possible to format a numeric based on values of specific digits in a PROC FORMAT? For example, would it be possible to format these numbers based on the coloured digits only without a data step to split them out first? (for example, the thousands values, irrespective of what the other values are... is there a wildcard type approach?) 501210 701304 I'm aware how to do this in a data step but, having just concatenated the variables for efficiency, I don't want to split them back out for formatting. EDIT: Adding context that I should have included in the first place! The format of the variable could be either numeric or character - either works for my purposes, by character is easier. The use of the format would for grouping in a class in a PROC TABULATE. Definitely not in a PUT statement as I don't want to split the data back out. For further context ... as an example, if my data has many variables in one row like: Value_at_8am [2 digits], Value_at_9am [2 digits], Value_at_10am [2 digits] ... and I concatenate them, can I then have a format that classifies based on the Value_at_9am in the new concatenated string irrespective of the value that come before or after it in the concatenated string? So, Ideally, I have a concatenated string when any range 01-99 in the 3rd and 4th places could be formatted to "True": Proc Format Value wishful **01** - **99** = "Value at 9am is True" **00** = "Value at 9am is False" ****01-****99 = "Value at 10 am is True" ... etc. cheers.
... View more
