Thanks! I run into the error with the code (when trying your second piece of code), saying that:   ERROR: _sentinel2 does not follow _sentinel1 on the list of previously defined variables.     I'd assume I would change _sentinel2 to just sentinel2 right? to this:   data need2 (drop=_sentinel:) /view=need2; set need1; by _sentinel1--sentinel2 notsorted; if first.sentinel2; run;   Also, if I attempt to run the above code, it just creates a NEED2 with no filtered data. What should I do to fix this? ... View more

Great! Thanks for your help. I made a new data set called Work.Delinquentloans and used your code to narrow most of it down:   data WORK.DELINQUENTLOANS; set work.data; where PERIOD_END_LSTAT = "In Grace Period" or PERIOD_END_LSTAT = "Late (31-120 days)" or PERIOD_END_LSTAT = "Late (16-30 days)"; run; data need1 / view=need1; set WORK.DELINQUENTLOANS; retain _sentinel .; run; data need2 (drop=_sentinel:) /view=need2; set need1; by _all_ notsorted; if first._sentinel; run; proc print data=need2; run; The last thing I realized running the new code is that there are a couple variables (INT_PAID, PRINCP_PAID and RECIEVED_AMT) that have values that need to be ignored for searching for duplicates. As you can see in Obs 3,4 & 5, it's the same loan but those specific variables are making it think it's three different loans. Any ideas on how to ignore some variables? Thanks again, it's much appreciated!   ... View more

Thanks for your reply!   So in your code, if the main data set is titled "work.data", would I be replacing the "have" in data need1 with "work.data"? or do I replace need1 with work.data?     data need1 / view=need1; set work.data; retain _sentinel .; run; data need2 (drop=_sentinel:) /view=need2; set need1; by _all_; if first._sentinel; run; proc print data=need2; run; or does it need to look like this?   data work.data / view=work.data; set have; retain _sentinel .; run; data need2 (drop=_sentinel:) /view=need2; set work.data; by _all_; if first._sentinel; run; proc print data=need2; run; ... View more

Well my main goal right now is to just make the table removed of all the duplicates where all of the columns are the same to the one above it (as in obs 339 and 340). I would want it to display all of the columns as shown in the screenshot. I mean the output would just need to be the removed duplicates table and a count of how many loans there are. The minimum list of variables I have decided to identify a loan are ID, IssuedDate and loan_amnt (not shown on screenshot): proc sort data=DATA out=SAS_Sorted_DATA nodupkey; by id loan_amnt IssuedDate; run;  Hopefully that helps!  ... View more