How to subset first row of each group?

[Birk Haugan]

I have a dataframe of about 3 000 acquisition bids.

Variables:

• Company ID (PERMNO)
• Bid ID (ID)
• Announcement date (ANDATE)
• Bidding round (ROUND_ID)

Bids are in competition with each other and grouped by ROUND_ID.

Problem:
I would like to write a code that runs through each ROUND_ID and saves the first observation (by ANDATE) in a separate dataframe. I.e. I would like to subset the first bid of each bidding round, and then create a different subset with the follower-bids.

Can anyone help me with this?

[Joro Kolev]

I am on old Stata 15 which does not have frames, but what you seem to want is

bysort ROUND_ID (ANDATE): keep if _n==1

If you still cannot figure it out, provide a usable sample of your data using -dataex-.