BUG: Weird column name type conversion in read_json with orient='table' and string columns · Issue #25435 · pandas-dev/pandas (original) (raw)

Code Sample, a copy-pastable example if possible

expected = pd.DataFrame({'1.': [1, 2]}) result = pd.read_json(expected.to_json(orient='table'), orient='table')

In [7]: expected                                                                                                                               
Out[7]: 
   1.
0   1
1   2

In [8]: result                                                                                                                                 
Out[8]: 
   1970-01-01 00:00:01
0                    1
1                    2

Problem description

There is a weird column type conversion from string to Timestamp.

Expected Output

The result should be equal to expected.

Output of pd.show_versions()

AttributeError: module 'tables' has no attribute 'version'