@ballardw wrote: @aanan1417 wrote: I HAVE A TABLE THAT CONTAIN DATA OF THREE  YEAR APPENDED DAY WISE . I HAVETO CHECK WHETHER NEW COLUMN ADDED OR NOT. THE DATA THAT I HAVE IS OF THREE YEAR "Appended" how? Is an existing text file replaced with a new one? A different text file every day? Are you replacing an existing SAS data set in the process (if so, stop doing that so you have the previous one to compare with). What is the result of this operation supposed to look like? A report when a "new column" is added? What would that report look like? Or do you want a change data set of some sort to reference?   This is a very broad subject and requires lots of details.   I once upon a time had a job organizing data from a number of collection sites. Since the purpose of the data was partially testing instrumentation data was added and removed from the data files often. In the 250K lines of code maintained to read the 40+ instrument sites I had code that identified changes because I maintained a data base of which report channels and positions were in use. So the code could report when a new channel (record level identifier) or report item (column identifier) were encountered. But that code was very dependent on the type of data collection was performed and identified. Random text may not be amenable to such. And all that the identification of new data did was tell me when and what the identifiers were. I then had to go to the guys running the data collection to get details like what each channel meant and the values collected in each position. Then updated a data base that was used to write/update code to read the files into usable data sets.   Other people making unannounced changes to data you are expected to process is a traditional definition of job security embedded in a state of annoyance. ... View more

Are you trying to install SAS or use SAS on the cloud. "ODAMID" is part of the URL for the cloud, not sure if it's also used for downloads for purchased version. Can you please be more specific as to what your exact issue is? What are you trying to do and what is not working? ... View more

Another way to eliminate this copy step would be to modify the earlier step that creates test1.CHDR. Why not have that earlier step create sas0416.CHDR_202210 instead? ... View more

Have you got instructions how to create the output? what are the rules ? I suppose that if data on both datasets is the same then remark="OK"; else remark="NOTOK";   To achieve that use merge statement: data want(drop=_name); merge have(in=in1) have2(in=in2 rename=(name=_name)); by ID; if in1 and in2 and name=_name then remark='ok'; else if in1 and in2 then remark = 'notok'; else delete; run;   ... View more

Hi:   In order to do what you want, the PREMIUM variable will need to be numeric. You can't add or summarize PREMIUM if it stays as a character variable as you show in your posted code.   Also, it's not clear if you need a TABLE as output or if you just need a REPORT. If you just need a REPORT, then either PROC REPORT or PROC TABULATE or other procedures could generate the summarized information you show. I chose PROC REPORT because the structure of the output table created in the OUT= option will be the same structure in the output dataset.   The only tricky part is that PROC REPORT will use absolute column numbers _C3_ and _C4_ for the COVERAGE columns in the output dataset, which required a RENAME, as shown in the second examples below:   Hope this helps you with your question. Cynthia ... View more

You never (as in NEVER) want to store dates or times as character, as you can't work with them. Read and store them as SAS date and time values in the first place. data want; input date :date9.; format date yymmdd10.; /* use any date format which suits you here */ datalines; 6-Mar-22 ; Dates are counts of days, starting with 1960-01-01 as day zero, going back to 1582 (the introduction of the Gregorian calendar) and forward to at least 9999-12-31. Times are counts of seconds, starting at midnight, or 1960-012-01T00:00:00 for datetimes.   ... View more