In my experience, a many-to-many merge in SAS almost always is a signal that there's data problems. Here, for example, age for ID=2 in the left-hand dataset is 10 or 20 ... well, which is it? And in the right-hand dataset, ID=2 has a salary of 5000 ... or 6000 or 7000. At any given point in time, person-ID=2's salary has to be one value. So probably/possibly, the right dataset has a date field that should have been used as a filter of some sort, and the programmer didn't know that, or didn't do it. I always interpret a SAS-note warning of a many-to-many merge as saying "Whoa Dummy ... you don't know your data as well as you think you do ... something needs fixing here." Offhand it's hard to envision a situation where what happens in a SAS many-to-many merge is what you WANT to happen, particularly in regard to how it carries (retains) field values into the "extra rows". But other than that, I strongly prefer SAS merges to SQL joins. You have clear control with the IN= vars, and the notes telling you the input and output counts of rows and fields is a tremendous help in auditing that what's happening is what you intended and expected.
Thank you for sharing. It was very helpful to clear the merge concept with same variables. Faced this with three table merge and when I was trying to do same step in Python didn't happened. So SAS was very handy to get the data with merge.
Sir thank you for sharing but to do proper merging or joining we need as first explained id, age ,salery 3 records but not should not show the age proper output
I have never seen the teaching like this, amazing, fabulous, elegent, fantastic, and Thank you so much for sharing your knowledge, I would like to subscribe to your channel one million times, plz go ahead
