BookmarkSubscribeRSS Feed
chatty
Calcite | Level 5

I am trying to merge a txt file and a SAS file. The text file (so tried to change it via dbms=dlm) and the file is not configured properly and I have to assign variables to certain columns (eg columns 6-7 represent age, but we have to code for that bc the numbers are just raw data). The names in the SAS file are too long and so I tried renaming them something more code appropriate eg weight_lbs but also received errors. When trying to do a "keep" command to show the variables to keep, I got the error  for both. the sets are called "dep" and "nutrition" error is "work.dep.data does not exist" and "work.nutrition.data does not exist". Bc of this wen I try to merge it says that there are no shared values, but I want to the values of only observations in both files. 

2 REPLIES 2
tomrvincent
Rhodochrosite | Level 12
I'd suggest working with each one individually. Get familiar with the data in each and then make enhancements to get them to merge.
Kurt_Bremser
Super User

What do you mean by "SAS file"? Is it a SAS dataset?

 

For detailed help, we need to know the data you are working with.

Post a few lines of the text file (including the header line, if there is such), using the {i} button, so the data is posted as-is.

For SAS datasets, post code with a data step with datalines, so we can create an exact copy of your dataset. Use the "little running man" for posting code.

A macro to automatically convert SAS datasets to data step code can be found in my footnotes.