BUG: use_inf_as_na options raises value error for read_csv · Issue #35493 · pandas-dev/pandas (original) (raw)

import pandas as pd import numpy as np pd.set_option('use_inf_as_na', True) pd.DataFrame({'test_data':[1,3,4,np.nan]}).to_csv('test_data.csv', na_rep='NaN') pd.read_csv('test_data.csv',sep=',' ,na_values='NaN')

Traceback (most recent call last): File "", line 1, in File "/home/lena/anaconda3/envs/pandas_pip2/lib/python3.8/site-packages/pandas/io/parsers.py", line 686, in read_csv return _read(filepath_or_buffer, kwds) File "/home/lena/anaconda3/envs/pandas_pip2/lib/python3.8/site-packages/pandas/io/parsers.py", line 458, in _read data = parser.read(nrows) File "/home/lena/anaconda3/envs/pandas_pip2/lib/python3.8/site-packages/pandas/io/parsers.py", line 1201, in read df = DataFrame(col_dict, columns=columns, index=index) File "/home/lena/anaconda3/envs/pandas_pip2/lib/python3.8/site-packages/pandas/core/frame.py", line 467, in init mgr = init_dict(data, index, columns, dtype=dtype) File "/home/lena/anaconda3/envs/pandas_pip2/lib/python3.8/site-packages/pandas/core/internals/construction.py", line 250, in init_dict missing = arrays.isna() File "/home/lena/anaconda3/envs/pandas_pip2/lib/python3.8/site-packages/pandas/core/series.py", line 4795, in isna return super().isna() File "/home/lena/anaconda3/envs/pandas_pip2/lib/python3.8/site-packages/pandas/core/generic.py", line 7109, in isna return isna(self).finalize(self, method="isna") File "/home/lena/anaconda3/envs/pandas_pip2/lib/python3.8/site-packages/pandas/core/dtypes/missing.py", line 124, in isna return _isna(obj) File "/home/lena/anaconda3/envs/pandas_pip2/lib/python3.8/site-packages/pandas/core/dtypes/missing.py", line 157, in _isna return _isna_ndarraylike(obj, inf_as_na=inf_as_na) File "/home/lena/anaconda3/envs/pandas_pip2/lib/python3.8/site-packages/pandas/core/dtypes/missing.py", line 218, in _isna_ndarraylike result = _isna_string_dtype(values, dtype, inf_as_na=inf_as_na) File "/home/lena/anaconda3/envs/pandas_pip2/lib/python3.8/site-packages/pandas/core/dtypes/missing.py", line 246, in _isna_string_dtype vec = libmissing.isnaobj_old(values.ravel()) File "pandas/_libs/missing.pyx", line 160, in pandas._libs.missing.isnaobj_old ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

Nans are not recognized as expected anymore if pandas option use_inf_as_nais set to True. Occurred first after upgrading to pandas 1.1.0