How does Pandas Merge as of (pd.merge_asof) 'left_by' and right_by' work?

Asked Jul 01 '19 at 14:13

Active Jul 01 '19 at 14:18

Viewed 5,169 times

I have a dataframe with two columns in which one is 'date_time' value and other (position) is a float. I have to merge two dataframes such that difference between the date_time values of the two frames is less than 30 min and position difference is less than 100 (metres). How do I do this?. I used merge_asof and used left_by and right_by.But its giving the error

MergeError: No common columns to perform merge on. Merge options: left_on=None, right_on=None, left_index=False, right_index=False

I have a common column 'ID' though it doesnt have same values in both frames. How can I accomplish this.From Documentation my understanding is that left_by and right_by is used when more than two columns is used.I have also tolerance to get the difference between them ~ 30 min ,but I also need the position difference to be ~ 100m Can I give tolerance as a list like ('tolerance' = pd.Timedelta('30min'),100) I dont know. Or if not possible is there some other way?

edited Jul 01 '19 at 14:18

asked Jul 01 '19 at 14:13

Fasty

5

take a little bit time reading https://stackoverflow.com/questions/53645882/pandas-merging-101/53645883 – BENY Jul 01 '19 at 14:15
There is no 'how' option in pd.merge_asof! – Fasty Jul 01 '19 at 14:16
I am using pd.merge_asof – Fasty Jul 01 '19 at 14:18
by is for exactly merge , on is for the tolerance merge in merge_asof – BENY Jul 01 '19 at 14:20
Ok So can I give multiple columns for merge and only merge them if both column tolerance is as mentioned above – Fasty Jul 01 '19 at 14:22
Multiple will not work – BENY Jul 01 '19 at 14:24

How does Pandas Merge as of (pd.merge_asof) 'left_by' and right_by' work?

0 Answers0