Update: The first dataset is using a previous SAS code that imports data from Excel. The excel file, has the PIDs of AA-1000, but in this code is where it gets truncated to AA-100 when we run the code.    So now we believe the issue to come from this initial setup SAS code.   Or is this an issue with Excel?    When we run the setup program we have a lot of import statements for multiple Excel files and sheets. When we merge the data, we get the same errors I had shown before, about various lengths being given to the PID variable as well as others.   Running PROC Contents after merging shows that PID still has a format and informat of $6.  I also ran PROC Contents on the first dataset imported from Excel and it gives me the same results.   We are a little stumped because this is the very beginning of the data analysis. I even asked to be sure if there was nothing that came before that might have defined PID from a previous program but no. We have a format catalog inside one of the libraries we are referencing so I thought maybe the issue could be there but from what they told me there is nothing before this program that defined PID with something from the format catalog.    ... View more

I have it 😁 ... View more

Okay, update: I got some of the PIDS from 999 to show up. Like for example, 1001, 1004, 1006, etc.   The issue wasn't really the informat and formats on the SAS code I believe. We are exporting from Access into the first database. This database has patients like 1002, 1005, 1007, 1011.    When it comes to the export of these variables from Access, this is what causes the truncation of data. 1002 gets cut to 100, and etc.  So it seems like there is an issue with either Access or the initial SAS program that imports all of the data from Access. Will work on it soon, thank you everyone for all of the help. While I don't have a solution yet, I feel like I am making small breakthroughs as a SAS rookie.  ... View more

Here's the first proc contents: And the second one:     ... View more

Is MERGE case sensitive when it comes to variables? Will PID in one dataset match with pid from the other one? ... View more

I did it to both data sets to set the same length for both before merging them ... View more

I did try this using both datasets just to make sure there was no further discrepancies but I got the same issue. Once it gets to AA-1000 it shortens it to AA-100 ... View more

Just got clarification on this.  I asked my colleagues and this is what they told me:   The software used to capture the data exports SAS code alongside a CSV file with the data. Running the SAS program in the same folder as the CSV file generates the dataset.      ... View more

Okay, there might be an issue in the way the import of the csv was handled. The sas code I was given to generate the dataset does not seem to have a proc import statement.   It just reads: %let csv_file = 'file_path' OPTIONS nofmterr; Then there are just numerous proc format statements for the variable informats and formats and labels and values.   ... View more

Yeah sorry so the variable that we use to identify has a prefix usually it's a two letter code like AR and a dash character . So for example, the full PID would be "AA-001."   When we start getting to values such as "AA-1000" this is where the value gets cut off and displays "AA-100" and so on.     ... View more

Both datsets do have the variable I am merging by in common. The only discrepancy seems to be the length in one of the datasets but I am not sure if it has something to do with the SAS code or the csv where I am pulling the data from.  ... View more

Anyone know how I can use this and not output all of the FREQ of the variables? Thank you! ... View more

This is super useful, and I have used it already. I am just trying to do some work myself so I can learn how SAS works. Macros are still fairly new, but I appreciate all of the help! You mentioned previously I could add a format correct? Can I save that as a separate .sas program and include it in this one? ... View more

Oh! This was totally the issue. I had commented out the "by variableName" by mistake ... View more