merge_asof() returns cryptic error with bad argument types

[sdfordham]

merge_asof arguments have to "...be a numeric column, such as datetimelike, integer, or float" but gives cryptic TypeError instead of ValueError on bad argument types?

import pandas as pd

left = pd.Series(['2019-06-01 00:09:12', '2019-06-01 00:10:29',
       '2019-06-01 00:20:30'], name='time')
right = pd.Series(['2019-06-01 00:01:12', '2019-06-01 00:01:12',
       '2019-06-01 00:01:12'], name='time')
pd.merge_asof(left, right)

Output:

...
TypeError: No matching signature found

versus e.g.

import pandas as pd

left = pd.Series(['2019-06-01 00:09:12', '2019-06-01 00:10:29',
       '2019-06-01 00:20:30'], name='time')
right = pd.Series(['2019-06-01 00:01:12', '2019-06-01 00:01:12',
       '2019-06-01 00:01:12'], name='time')

left = pd.to_datetime(left, infer_datetime_format=True)
right = pd.to_datetime(right, infer_datetime_format=True)

pd.merge_asof(left, right)

[TomAugspurger]

A better error message would be welcome here. I'm not sure where is best, but probably somewhere in _get_join_indexers in core/reshape/merge.py

[gabriellm1]

Can I help it? We just need to raise a ValueError in case the arguments in _get_join_indexers are not datetimelike, float or integer?

[n0nameNPC]

merge_asof still doesnot work

TypeError                                 Traceback (most recent call last)
<ipython-input-397-5864196bc3ac> in <module>
      3 right = pd.Series(['2019-06-01 00:01:12', '2019-06-01 00:01:12',
      4        '2019-06-01 00:01:12'], name='time')
----> 5 pd.merge_asof(left, right)

~\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.9_qbz5n2kfra8p0\LocalCache\local-packages\Python39\site-packages\pandas\core\reshape\merge.py in merge_asof(left, right, on, left_on, right_on, left_index, right_index, by, left_by, right_by, suffixes, tolerance, allow_exact_matches, direction)
    566         direction=direction,
    567     )
--> 568     return op.get_result()
    569 
    570 

~\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.9_qbz5n2kfra8p0\LocalCache\local-packages\Python39\site-packages\pandas\core\reshape\merge.py in get_result(self)
   1572 
   1573     def get_result(self):
-> 1574         join_index, left_indexer, right_indexer = self._get_join_info()
   1575 
   1576         llabels, rlabels = _items_overlap_with_suffix(

~\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.9_qbz5n2kfra8p0\LocalCache\local-packages\Python39\site-packages\pandas\core\reshape\merge.py in _get_join_info(self)
    907             )
    908         else:
--> 909             (left_indexer, right_indexer) = self._get_join_indexers()
    910 
    911             if self.right_index:

~\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.9_qbz5n2kfra8p0\LocalCache\local-packages\Python39\site-packages\pandas\core\reshape\merge.py in _get_join_indexers(self)
   1878             # choose appropriate function by type
   1879             func = _asof_function(self.direction)
-> 1880             return func(left_values, right_values, self.allow_exact_matches, tolerance)
   1881 
   1882 

pandas\_libs\join.pyx in pandas._libs.join.__pyx_fused_cpdef()

TypeError: No matching signature found

now, working with these constructions:

success = confirmed_ts_p_day.set_index('Timestamp').reindex(df.set_index('Timestamp').index, method='nearest').reset_index()
new_df = pd.merge(df,success, on='Timestamp')

[MarcoGorelli]

@Aporbel please open a new issue with a reproducible example