AttributeError: module 'pandas.core.ops'

[mboldin]

I am running on Windows 10 Intel i5
and get import error with

import pandas as pd
import text_extensions_for_pandas as tp
when run in Jupyter Notebook

the relevant error code is
AttributeError: module 'pandas.core.ops' has no attribute '_get_op_name'

imports worked 6 weeks ago fail now and
I have no problem with import text_extensions_for_pandas not in Jupyter
I tried to create a virtenv just for this problem and installed full pip install of pandas, text_extensions_for_pandas and jupyter
and I see the same error only in Jupyter
see ERROR in context for more detail

Versions
pandas 1.3.5

and
Selected Jupyter core packages...
IPython : 7.34.0
ipykernel : 6.15.2
ipywidgets : 8.0.2
jupyter_client : 7.3.5
jupyter_core : 4.11.1
jupyter_server : not installed
jupyterlab : not installed
nbclient : 0.6.7
nbconvert : 7.0.0
nbformat : 5.4.0
notebook : 6.4.12
qtconsole : 5.3.2
traitlets : 5.3.0

I am not 100% sure of the text_extensions_for_pandas versions maybe 0.2.0

(text_pandas) C:\Users\mdbol\Desktop\virtenv\text_pandas>pip install text_extensions_for_pandas
Collecting text_extensions_for_pandas
Using cached text_extensions_for_pandas-0.2.0-py3-none-any.whl (199 kB)

---

ERROR In context:
~\AppData\Local\Programs\Python\Python37\lib\site-packages\text_extensions_for_pandas\array_init_.py in
25 from text_extensions_for_pandas.array.char_span import *
26 from text_extensions_for_pandas.array.token_span import *
---> 27 from text_extensions_for_pandas.array.tensor import *
28 from text_extensions_for_pandas.array.accessor import *
29

~\AppData\Local\Programs\Python\Python37\lib\site-packages\text_extensions_for_pandas\array\tensor.py in
281
282 # Add operators from the mixin to the class
--> 283 TensorArray._add_arithmetic_ops()
284 TensorArray._add_comparison_ops()

~\AppData\Local\Programs\Python\Python37\lib\site-packages\pandas\core\arrays\base.py in _add_arithmetic_ops(cls)
1342 @classmethod
1343 def _add_arithmetic_ops(cls):
-> 1344 setattr(cls, "add", cls._create_arithmetic_method(operator.add))
1345 setattr(cls, "radd", cls._create_arithmetic_method(ops.radd))
1346 setattr(cls, "sub", cls._create_arithmetic_method(operator.sub))

~\AppData\Local\Programs\Python\Python37\lib\site-packages\pandas\core\arrays\base.py in _create_arithmetic_method(cls, op)
1495 @classmethod
1496 def _create_arithmetic_method(cls, op):
-> 1497 return cls._create_method(op)
1498
1499 @classmethod

~\AppData\Local\Programs\Python\Python37\lib\site-packages\text_extensions_for_pandas\array\tensor.py in _create_method(cls, op, coerce_to_dtype)
92 return TensorArray(res)
93
---> 94 op_name = ops._get_op_name(op, True)
95 return set_function_name(_binop, op_name, cls)
96

AttributeError: module 'pandas.core.ops' has no attribute '_get_op_name'

[mboldin]

UPDATE:

Looking at the Jupyter consoile output, I see:

[IPKernelApp] ERROR | No such comm target registered: jupyter.widget.control
[IPKernelApp] WARNING | No such comm: 86364858-6645-43b9-8ee3-3f3405095abe

when import text_extensions_for_pandas fails

A Google search of this error string shows it has been reported for other packages (not text_extensions_for_pandas)
as a result of problems related to ipywidgets see jupyter-widgets/ipywidgets#2257

So I doubt it is a fixable TEXT_EXTENSIONS_FOR_PANDAS
I will close this issue in a few days if this diagnosis is right, and in meantime hope someone report here a workaround.

[mboldin]

FIXED -- for me
My jupyter run was pointing to an old pre-alpha version of text_extensions_for_pandas
I deleted this, and forced a new install of he module and all is OK
The bad line
---> 94 op_name = ops._get_op_name(op, True)
was from an old version of tensor.py
the latest 0.2.0 version has different code logic for
def _create_method(cls, op, coerce_to_dtype=True, result_dtype=None):

[frreiss]

Glad this issue is resolved for you now. Let us know if you have any other problems.