Merging two datasets

[Fahmida Huq]

Hello,

I need to merge two datasets. I already have one existing dta file, and I converted one csv file to dta,and tried to merge it with the existing dta fie. But whenever I'm trying to do it, it says "merge already defined". Can someone help me out to solve this problem? Thanks.

Fahmida

[Carole J. Wilson]

Whenever you merge a file, Stata creates a new variable "_merge". Does this variable already exist in your dataset? If so, drop/rename that variable before attempting to merge (or use the gen() option to choose a different name for the variable Stata will create in the merge).

[Fahmida Huq]

Thanks. I have got it. However, if I am merging two datasets using a variable, how to make it uniquely identify observations in the existing file? For example, when I merged two datasets by a variable name hsn, it is saying-

(note: variable hsn was str4, now str25 to accommodate using data's values)
variable hsn does not uniquely identify observations in the using data

My question is- how can I make it compatible to identify observations?

Fahmida

[Rich Goldstein]

the note is not a problem; that hsn is not unique is a problem; you can use the -duplicates- command to find the problems; see

help duplicates

there are other ways also; e.g.,

sort hsn
browse hsn if hsn==hsn[_n-1]