Either tell PROC IMPORT that it should check more of the file before deciding for you what the variables are. https://support.sas.com/documentation/cdl/en/proc/70377/HTML/default/viewer.htm#n1qn5sclnu2l9dn1w61ifw8wqhts.htm Just add this statement into your PROC IMPORT code. guessingrows=max; Or better still you could just read the file yourself instead of letting PROC IMPORT guess at what is in the file.  You are probably going to be able to make a better decision about what each column represents than SAS can guess.  You will understand the meaning of the columne and SAS can only look at the single example input file you have given it.  So you might have a sample file with really short strings, but you would know to define the variables longer to handle other data files of the same type.   It is really easy to write a data step to read a delimited file. Follow these simple steps. First write a DATA statement and an INFILE statement.  Use the DSD option to indicate that you want to properly handle missing values. The default delimiter for DSD is a comma, but you can use the DLM= option to change that. Use the FIRSOBS=2 option to skip the header line (if your file has one).  Use the TRUNCOVER option in case some lines do not have all of the columns. data want ; infile 'myfile.csv' dsd dlm=',' truncover firstobs=2; Then copy the first line from your delimited file and paste it into your program and convert it to a LENGTH statement. This will let you can define the variable's position, name, type and length.  If the headers are not valid SAS variable names you will need to edit them to make them into valid names.  You could even take this opportunity to convert the column headers into variable names that might work better for you.  Lengths with $ in front mean the variable is character. Otherwise use 8 for the length as SAS stores all numbers (dates and times are numbers also) as 64bit floating point numbers. length fdate 8 mgrname $40 Country $20 mgrno $10 typecode $10 rdate 8 prdate 8; Then for any variables that REQUIRE them add INFORMAT and/or FORMAT statement. So if your "date" variables are actual dates then you will probably want to use an INFORMAT and a FORMAT so that SAS will treat them as dates and not numbers or strings. There is no need to apply informat or formats to character variables or even most numeric variables.  There is no need to add a length to the INFORMAT since SAS will ignore it anyway when using the DSD option on the INFILE statement. informat fdate rdate prdate yymmdd.; format fdate rdate prdate yymmdd10.; Then for the input statement you just need to list the variables.  If you have defined them in the right order in the LENGTH statement then you can use a variable list in the INPUT statement to make it even easier. input fdate -- prdate ; run; So the PROC IMPORT program will be 5 statements over 7 lines and the DATA step program is 7 statements over 7 lines.  5 statements if you just want to read the dates as strings. ... View more

"On average there are 3 observations in the same group. But that is exactly what I need it to be to get my result."   And you want 1 regression per by group?   Models won't even be estimated if the by group size does not have 2 more obs than there are independent variiables.  What  possible use would it be? ... View more

Use INTNX and NWKDOM function.  ... View more

@Haikuo wrote: @Kurt_Bremser, If the new_column is setup as length of 4, then SUBSTR() seems redundant in this context. Until, for some reason, you need your target variable to be longer to also hold values from somewhere else, but the requirement for the first 4 digits HERE did not change. Then you suddenly get more than 4 characters, and your results go bonkers just because you wanted to avoid some typing. Keep in mind that the length statement might be pages away from the place where you use the variable(s). Always make your code explicit so the intention is clearly visible right on the spot.   Been there, done that, and had to spend days to debug it. ... View more

Maybe start by defining what your problem is.  What is an event study in your context, what is the problem you are having with it.  Remember this is a question and answer forum, not a contract work forum. ... View more

Thank you. The following worked. In fact, I was trying to retrieve the Monday after the third Saturday of every month. IF (DATES=nwkdom(3, 7, month, year)+2); RUN; ... View more