I have df1:
a
2013
2014
2015
and df2
a
2014
2017
2018
2021
My goal would be that the key a from df1 merges onto the next highest key in df2 e.g. 2013->2014, 2014->2017 and 2015->2017. So even if both keys are equal I'd like to merge on the next highest, would that be possible?
Use merge_asof with direction='forward' and parameter allow_exact_matches=False:
df = pd.merge_asof(df1, df2.rename(columns={'a':'a1'}),
left_on='a',
right_on='a1',
direction='forward',
allow_exact_matches=False)
print (df)
a a1
0 2013 2014
1 2014 2017
2 2015 2017