Hi @Quentin ! I actually ended up running the code three times (4 corvars at a time) since the computer kept running out of disk space. If I recall correctly, it was 20 billion or so for each run. ... View more

Thank you so much for this code, @Tom ! I will explore it further. ... View more

You are absolutely right, @PaigeMiller.  Next time If I have another question, I will make sure to include all the important details. I sincerely apologize for this. Thank you so much for all your help and the code that your proposed. It worked like a charm. ... View more

Great question, @PaigeMiller ! This is related to the need to show that the results are robust to different ways of measuring the variable of interest and to the different subsamples. Most of them will be a footnote anyways but we need to perform them. ... View more

Thank you so much for sharing this other perspective, @Reeza . I need to think about how it will work with the data. The starting dataset has around 50 million rows of data (another one that I need to apply the same code to has about 250 million).  ... View more

Quick clarification: by "0 or 1" and "0,1" I meant that we include all the values where the variable is either 0 or 1, which by definition is all the observations for a given dummy variable. ... View more

Thank you @Reeza for this question. Your statement is correct. The combinations will 0, 1, and 0&1 together versus 0, 1, and 0&1 together. I have 12 variables (corvar1 through corvar12), and each corvar takes the values of 0 and 1. And the other dummy, allmdummy, takes the values of 0 and 1 too. I want to be able to run the same regression for every single combination, which amounts to 108. All the possible combinations for the first variable will look like the following: corvar1=0 and allmdummy=0,   corvar1=0 and  allmdummy=1 corvar1=0 and allmdummy= 0 or 1   corvar1=1 and allmdummy=0, corvar1=1 and allmdummy=1 corvar1=1 and allmdummy= 0 or 1   corvar1=0,1, allmdummy=0, corvar1=0,1 allmdummy=1 corvar1=0,1 allmdummy= 0 or 1   Ideally, I will be able to create a unique sampleid for each combination and then run proc reg and proc means command for each combination (sampleid). I hope this helps. Please let me know if you have any follow-up questions. Thank you so much! ... View more

Thank you so much for your suggestions, @Quentin . If I could find a way to do the proc reg without splitting the dataset, that will actually be awesome since the splitting and re-combining will take a massive amount of hard disk space. The challenge I have at this point is that some observations will fall under more than one sample ID due to overlapping subsets. ... View more

Thank you so much once again for your quickly reply, @PaigeMiller . Here is my challenge: the subsets will be overlapping. For example, sampleid 111 will be a combination of 112 and 113. I could not figure out how to create sample IDs without overwriting the existing ones. ... View more

Thank you so much, @PaigeMiller  for your prompt reply. You made an excellent point. It is a nice segway into my end goal. I actually want to run the following for each sampleid (assuming that wanted is the output of the prior macro). If I could find a way to incorporate the prior macro into the next one, that will indeed save me a lot of time. I will really appreciate any insight on this one.   proc reg data=wanted noprint outest=wanted2 noprint adjrsq; by sample id obsid;    /*obsid is another ID already stored in the dataset*/ model depvar1 = ind1 ind2 ind3 ind4 / hcc; run; proc means data=wanted2; by sampleid; var intercept ind1 ind2 ind3 ind4 _rsq_ _adjrsq_; output out=wanted3 mean = std= t= probt= / autoname; run; ... View more

Thank you so much @ballardw , @Patrick , and @mkeintz  for helping me with my question. I sincerely appreciated all your help. Have a wonderful weekend! ... View more

Thank you so much @Patrick for this suggestion. When I ran it, it did not quite produce the dataset I would like to have. day_count starts for the first observation in the sample and goes all the way till the end of the sample instead of starting at each secid & eventday=1 pairs. Should I create an ID that combines secid and event day to accomplish this task? Thank you so much ... View more

Thank you so much @ballardw  and @mkeintz for helping me with my inquiry. I really appreciated it. My data set extremely large, so perhaps the data step solution may work better for me. Please accept my apologies. The data set does have IDs, and I need to ensure that these steps are executed within the same IDs only.  I did not work with hash operators before. How should I go about bringing in the IDs? I attempted below (bold line), but ended up getting something different.   Also: I opted for 180 days instead of 6 months since I need to have 180 observations for each event, and the 6-month cutoff may not necessarily produce 180 observations. If for some events, I have fewer than 180 observations (due to data limitations), this will be fine. I will first run the estimations with only those with 180 observations first, and later I will include all of them (even those with fewer than 180 observations).   data have;; input secid obs_date :mmddyy10. event value; format obs_date mmddyy10.; datalines; 1 1/1/2020 0 0.003918297 1 1/2/2020 0 0.294439369 1 1/4/2020 1 0.048466828 1 1/5/2020 0 0.718177034 1 1/6/2020 0 0.001364211 1 1/7/2020 0 0.440971903 1 1/8/2020 0 0.55962246 1 1/9/2020 0 0.821105414 1 1/10/2020 1 0.332539126 1 1/11/2020 0 0.289808782 1 1/12/2020 0 0.911322784 1 1/18/2020 0 0.896737429 1 1/19/2020 0 0.333353538 1 1/20/2020 0 0.19605047 1 1/21/2020 0 0.60689124 1 1/22/2020 1 0.439833751 1 1/23/2020 0 0.896841702 1 1/24/2020 0 0.147458515 1 1/25/2020 0 0.923237859 1 1/26/2020 0 0.394894147 1 1/27/2020 0 0.935664763 1 1/28/2020 0 0.667981258 1 1/29/2020 1 0.872639458 1 1/30/2020 0 0.064826005 1 1/31/2020 0 0.109456422 1 2/1/2020 0 0.386348541 1 2/2/2020 0 0.732321068 1 2/3/2020 0 0.503449099 2 1/1/2020 0 0.003918297 2 1/2/2020 0 0.294439369 2 1/4/2020 1 0.048466828 2 1/5/2020 0 0.718177034 2 1/6/2020 0 0.001364211 2 1/7/2020 0 0.440971903 2 1/8/2020 0 0.55962246 2 1/9/2020 0 0.821105414 2 1/10/2020 1 0.332539126 2 1/11/2020 0 0.289808782 2 1/12/2020 0 0.911322784 2 1/18/2020 0 0.896737429 2 1/19/2020 0 0.333353538 2 1/20/2020 0 0.19605047 2 1/21/2020 0 0.60689124 2 1/22/2020 1 0.439833751 2 1/23/2020 0 0.896841702 2 1/24/2020 0 0.147458515 2 1/25/2020 0 0.923237859 2 1/26/2020 0 0.394894147 2 1/27/2020 0 0.935664763 2 1/28/2020 0 0.667981258 2 1/29/2020 1 0.872639458 2 1/30/2020 0 0.064826005 2 1/31/2020 0 0.109456422 2 2/1/2020 0 0.386348541 2 2/2/2020 0 0.732321068 2 2/3/2020 0 0.503449099 2 2/4/2020 0 0.60689124 2 2/5/2020 0 0.439833751 2 2/6/2020 0 0.896841702 2 2/7/2020 0 0.147458515 2 2/8/2020 0 0.923237859 ; data want (drop=event); set have (where=(event=1)); by secid; if _n_=1 then do; declare hash h (dataset:'have (drop=event)',ordered:'a'); h.definekey('obs_date'); h.definedata(all:'Y'); h.definedone(); declare hiter hi ('h'); h.output(dataset:'h'); end; day=0; eventday=obs_date; format eventday date9.; output; hi.setcur(); /*Point hash iterator at hash dataitem for event date */ do day=1 to 180 while (hi.next()=0); output; end; run;   ... View more