Tom, to your point, I may not need a loop.  However, I am trying to translate what I can do with VBA, where I create an array of all the files to process then loop through the array to process them all.  In the most simple terms, it would be something like this if I have an array called arFiles, with one element containing the file name, as my one dataset in SAS does For i = 1 to ubound(arFiles) step 1      process arFiles(i) Next i Now, in that, the 'process arFiles(i)' would take maybe 8 or 10 lines, but I hope you get the logic.  Perhaps this is not necessary in SAS, but my knowledge of how SAS processes is extremely limited, so I am constantly thinking in terms of how VBA would do something, and then I trial and error it until I can muddle a workable result.  So, in my dataset I have 65 filepaths to the Excel files I need to import.  My reasoning is that I use proc sql to get the first one that has not been processed, use proc import to import it, then move on to the next one until they are all done.  If you know of a good reference that would describe a better way to do it in one step I sure would appreciate it.  A lot of the posts I read here are just way beyond my comprehension of SAS, so I spend hours researching just the simple things like filevar and pipe.  Thank you. Greg ... View more

DOH!  I forgot double quotes around the macro variable in the update statement.  It now works.  Next step, put it all in a loop to step through all observations in the dataset. Greg ... View more

Thank you so much Tom, I knew it had to be something minor.  As for the last update statement, I'm still getting errors, but I think I can trial and error it until it works.  Thank you for your good time.  Thank all of you for that matter. Greg ... View more

Tom, thank you for the tip on the recursive dir command.  Indeed that would be easier, but I need to be able to filter out the folders and files I need, and I'm not sure how I would do that with your simpler method.  I'm stuck now in getting the proc import to work with a macro variable created by the file path stored in dataset flist.  My thought, and I need to keep it simple, is to set a macro variable with proc sql, then use that variable in proc import, but it is not working.  I have attached the file, and my problems start where I have put comments.  I sure would appreciate any tips, as at this point I've tried everything I understand how to do.  Thank you. Greg ... View more

Thank you both so much for your help so far.  As of now, I have all of the complete paths of the files I need in a dataset called 'FList'.  I'm a little bit stuck on how I should proceed to pass each file path to an import macro.  I thought it would just be a matter of passing each file path to an array, then loop through the array using each element as the file in a proc import statement, but I really can't get my head around it.  Once again, I am unable to paste into the editor, so I have attached what I have so far.  I realize I could probably simplify it, but for now I would like to keep it broken into steps until I fully understand what it is doing.  Thanks again. Greg ... View more

Huey, thank you for the reply.  I won't be able to try it until I get back to work on Monday, but I wanted to ask a question in the mean time.  I found this doc http://www2.sas.com/proceedings/sugi27/p082-27.pdf  and it only references ascii files.  The ultimate files I am trying to get to are .xlsx files.  Will your method work with Excel files?  From what you have posted, and what I have read, it seems like your suggestion is the way to go.  Thanks again. Greg ... View more

Ok, one more question regarging this.  I am looking at the "Analysis of Maximum Likelihood Estimates" output, and I am confused as to why some of the variables have '<.0001' in the 'Pr > ChiSq' column.  I would (maybe naively) infer these variables to be significant.  Am I wrong in this assumption?  It just seems to straight forward.  Thank you. Greg ... View more

Thank you Reeza, that is just what I needed.  I spent some time googling "sas logistic regression", and I came up with genmod, catmod, but not logistic.  Anyhow, thanks for the tip. Greg ... View more

Aad, by jove that is *exactly* what I wanted to do.  Yesterday I eventually settled on a couple of steps involving proc means, then deleteing all rows where _stat_ <> 'N'.  Anyhow, I found a document that explained what you have stated, and also the data= part.  Anyhow, thank you for the tip. Greg ... View more

Ok, I get it now.  Since I dropped all but the columns I want, I did not see there are multiple rows for each category, one for each statatistic.  It seems the only row I want to keep is where _stat_ = 'N'.  The missing values were the STD row, which makes total sense now.  ... View more

Well, that (what I posted above) can't be right either, since there are 36,925 obs in the want table, and only 27,519 in the source table.  I have no option here to use anything other than explorer, but thank you for the tip Linlin. ... View more

I think this will do it, but I'm not quite sure why there would be missing values for my 'N', which is just _freq_ renamed. proc means data=npso noprint;      by max_n cr_scor bad_debt_at_conn dep_unpaid chg_off;      output out=want (rename=(_freq_=N)                               keep=max_n cr_scor bad_debt_at_conn dep_unpaid chg_off n); run; Now, it seems my big problem is that I am unable to copy and paste anything in this window.  Does anyone else have that happen? Thank you Greg ... View more