Use of line comments such as in this line can have very unexpected results with macros: proc reg noprint data=&in_ds outest=oreg_ds1 edf; where date between &date1 and &date2; *Restricted to DATE1- DATE2 data range in the loop; Inside the body of a macro you should use:  %* comment text;   Or stick with the /* comment */ statement. ... View more

I guess you'd rather want to have this: %macro m(start=,end=); proc expand data=A1 (keep=ID date AAA) out=A3 method=none ; by permno; id date; %do i=&start %to &end; convert AAA = AAA_&i / transformout=(reverse MOVPROD &i reverse trimright %eval(&i-1)); %end; run; %mend m; %m(start=2,end=50); ... View more

@ZZB wrote: I am not sure if my sas has it. How to check it? If I have it, how to code it? I think that you may actually want IML, SAS's matrix language more than ETS. ETS does time series analysis but I haven't had a license for 20 years so no practice... To check license run: Proc setinit; run; If you see SAS/ETS in the output of the log you have a license for ETS.   If you have the license run: Proc product_status; run; to see if it is installed. If licensed and not installed then update the installation.   ... View more

1. Start by getting regression running for a single model 2. Expand your model and run it for three iterations.  3. Look at what's changed between 1 & 2 and that will tell you what you need to change for your model.    You can see some examples of this here: https://stats.idre.ucla.edu/sas/seminars/sas-macros-introduction/   And this is a commonly asked question, so if you search you should find results here or on lexjansen.com     ... View more

Hello ZZB; Have you ever solved this problem? I am facing the same issue. Could you help? please? Thanks ... View more

Better late than never?       convert high=high1 / transformin=(lag 1) transformout=(movmax 12 trimleft 10);   This makes the current value of high tied with the preceding, so max over 12 obs, is the same as max over preceding 11 obs.   Regards, Mark   ... View more

Try   proc expand data=XXX out=XXX method=none; by AAA; convert high = high1 / transformin=(movmax 11 lag 1 trimleft 10); run; ... View more

I would say that two endpoints don't really describe a desired axis very well. Do you want annual, month, quarterly or some other interval to appear on the axis? And is your date variable a SAS date or not? Some of the issue is that your are apparently trying to display 26 years of data. ... View more

Use function INTNX with "SAME" or "END" parameter : data have; input (inDate outDate) (:ddmmyy.); format inDate outDate ddmmyy10.; datalines; 31/10/2001   30/11/2001 30/11/2001   31/12/2001 31/12/2001   31/01/2002 31/01/2002   28/02/2002 15/02/2002   . ; data want; set have; format newDate endDate ddmmyy10.; newdate = intnx("MONTH", inDate, 1, "SAME"); enddate = intnx("MONTH", inDate, 1, "END"); run; proc print data=want noobs; run; PG ... View more

And just related to the appearance: Be very careful opening CSV files in Excel (and likely some other spreadsheets) as it tries to be "helpful" and treats what you thought was a character 0008899 as a numeric. If you save the file from Excel will change the actual value. ... View more