Hi all,
Could somebody help me with this issue?
I am a SAS student working on an assignment which is a small project about customer distribution and deactivation. I am importing the raw data into SAS with data step.
The last column is Sales. The length of sales values under this column vary from 5 to 7 in the format of $20.00, $105.12, and $1298.22. (Please see the attached file with a small fraction of the original data.) I want to import them into SAS in their original format including the dollar sign. After that, I will work on them for some analyses.
I tried different ways but all failed. I think the issue is that I don't know how to deal with the different lengths of values under that column.
I am looking forward to your help.
Thank you very much.
Sagum
You don't "import" from a text file. You just READ the text file. There are no "different lengths" under that column. SALES is always in columns 82 to 92 on every line for the file you shared.
Use the COMMA informat, just like you need to use the MMDDYY informat for the date fields. You can use the @ cursor motion command to position the cursor in the right place to read those strings that require informats.
data want;
infile 'c:\downloads\partial raw data.txt' truncover ;
input id $ 1-13 @15 date1 mmddyy10. @26 date2 mmddyy10. /* other variables */ @82 sales comma11. ;
format date1 date2 yymmdd10. sales dollar11.2;
run;
Results:
Obs id date1 date2 sales 1 1176913194483 1999-06-20 . $128.00 2 1176914599423 1999-10-04 1999-10-15 $72.00 3 1176951913656 2000-07-01 . $593.00 4 1176954000288 2000-05-30 . $83.00 5 1176969186303 2000-12-13 . . 6 1176991056273 1999-08-31 2000-09-18 $1,041.00 7 1176991866552 2000-05-24 . . 8 1176992889500 2000-11-28 . $72.00 9 1177000067271 1999-12-23 . $134.00 10 1177010940613 1999-12-09 . $11.00
You don't "import" from a text file. You just READ the text file. There are no "different lengths" under that column. SALES is always in columns 82 to 92 on every line for the file you shared.
Use the COMMA informat, just like you need to use the MMDDYY informat for the date fields. You can use the @ cursor motion command to position the cursor in the right place to read those strings that require informats.
data want;
infile 'c:\downloads\partial raw data.txt' truncover ;
input id $ 1-13 @15 date1 mmddyy10. @26 date2 mmddyy10. /* other variables */ @82 sales comma11. ;
format date1 date2 yymmdd10. sales dollar11.2;
run;
Results:
Obs id date1 date2 sales 1 1176913194483 1999-06-20 . $128.00 2 1176914599423 1999-10-04 1999-10-15 $72.00 3 1176951913656 2000-07-01 . $593.00 4 1176954000288 2000-05-30 . $83.00 5 1176969186303 2000-12-13 . . 6 1176991056273 1999-08-31 2000-09-18 $1,041.00 7 1176991866552 2000-05-24 . . 8 1176992889500 2000-11-28 . $72.00 9 1177000067271 1999-12-23 . $134.00 10 1177010940613 1999-12-09 . $11.00
Hi Tom,
Thank you so much for your prompt response. Your solution works.
There is a lot to learn for me.
Sagum
SAS Innovate 2025 is scheduled for May 6-9 in Orlando, FL. Sign up to be first to learn about the agenda and registration!
Learn how use the CAT functions in SAS to join values from multiple variables into a single value.
Find more tutorials on the SAS Users YouTube channel.
Ready to level-up your skills? Choose your own adventure.