topic Re: How to remove commas and dots from numerics in SAS Programming

How to remove commas and dots from numerics

sasowiczka123 — Mon, 18 May 2020 19:51:31 GMT

Hello,
During analyzing financial data I came across a huge difficulty. Variable "total salary" looks like this:

75,879.00 or 123,679.00 or 7,987.00. Do you know how to delete commas and periods? They occure in different places.

Please help

Re: How to remove commas and dots from numerics

Kurt_Bremser — Mon, 18 May 2020 20:00:38 GMT

If you want the numbers to display differently, assign another format. But mind that the display format does not influence any calculations.

Re: How to remove commas and dots from numerics

Reeza — Mon, 18 May 2020 20:14:01 GMT

What is the type and format applied on that variable? Is this after you've imported the data or something you're trying to do after you've imported the data?

Re: How to remove commas and dots from numerics

ballardw — Mon, 18 May 2020 20:14:46 GMT

Please describe what you are doing that the presence of either the comma or the decimal affects your analysis.

You may also want to run proc contents on that data set and share the variable type and format currently associated with the variable.

I suspect that the current format is some form of CommaW.D. The presence of the commas or decimals are very unlikely to have any impact on analysis though possible will have an effect on some reports.

Re: How to remove commas and dots from numerics

Shmuel — Mon, 18 May 2020 20:21:05 GMT

What is the origin of your data? is it excel or any other external file or data base?

How do you import that file into sas?

I believe that the issue is the result of method importing the data to sas.

Re: How to remove commas and dots from numerics

biopharma — Mon, 18 May 2020 20:33:35 GMT

What is the variable type? Numeric that is formatted to display such or a character variable containing the above strings? Don't think you want to remove the period without also removing the zeroes that follow. Otherwise, it affects the value.

If it is a numeric variable type you can clear the format with:

format <variable> ;

If it is character, you can compress the comma out. I don't think you should delete the period as it would change the value. Here is a small demo:

900        data _null_ ;

901          x = "123,679.00" ;

902          y = compress(scan(x,1,".")) ;

903          put x = y = ;

904        run ;

x=123,679.00 y=123,679