merge_asof question

unread,

Dec 9, 2016, 3:40:10 PM12/9/16

to PyData

Hi,

Is there a function similar to merge_asof that selects the first row in the right DF whose 'on' key is greater than (or equal to) the left key?

Thank you,

Peter

unread,

Dec 14, 2016, 2:26:04 PM12/14/16

to PyData

Off the top of my head, you could get the indices you want with

np.searchsorted(right['timestamp'].values, left['timestamp'].values)

This will return one-past-the-end for non-matching timestamps, so you can't directly use these indices with .iloc[] if there would be missing rows.

unread,

Dec 15, 2016, 11:06:55 AM12/15/16

to PyData

I've just added this as a new request:

unread,

Dec 19, 2016, 9:31:00 AM12/19/16

to PyData

search sorted has been useful, but definitely looking forward to the merge_asof update. Thank you for adding it as a new request Christopher.

Peter

Reply all

Reply to author

Forward