Skip to content

BUG: to_hdf failing with integer columns and data_columns=True #9057

Closed
@jreback

Description

@jreback
In [39]: N = 1000

In [40]: df = DataFrame(np.random.randint(0,8,size=N*400).astype('uint8').reshape(N,-1),index=pd.date_range('20130206',periods=N,freq='ms'))

In [41]: df.to_hdf('test.h5','df',mode='w',format='table')

In [42]: df.to_hdf('test.h5','df',mode='w',format='table',data_columns=True)
AttributeError: 'numpy.int64' object has no attribute 'startswith'

Metadata

Metadata

Assignees

No one assigned

    Labels

    BugDtype ConversionsUnexpected or buggy dtype conversionsIO HDF5read_hdf, HDFStore

    Type

    No type

    Projects

    No projects

    Milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions