str'ified column names in DataFrame.to_hdf(); Fixes #9057 #9902
Conversation
Converted column names to strings before creating pytables metadata object
| periods=N,freq='ms')) | ||
|
|
||
| self.filename = NamedTemporaryFile(suffix='.h5', delete=False).name | ||
| df.to_hdf(self.filename,'df', mode='w', format='table', |
There was a problem hiding this comment.
these all need to go in io/tests/test_pytables.py
|
Yeah, my bad, that makes sense. I moved them over. |
| @@ -4623,6 +4624,27 @@ def _test_sort(obj): | |||
| else: | |||
| raise ValueError('type not supported here') | |||
|
|
|||
| class TestToHdfWithIntegerColumnNames(tm.TestCase): | |||
| # GH9057 | |||
| def setUp(self): | |||
There was a problem hiding this comment.
pls test the way everything else is done
eg use ensure_clean for a file
There was a problem hiding this comment.
pls give a test with Unicode column names
|
the reason for not allowing this is that queries become non-obvious and the data does not correctly round trip so I suppoose this is ok but needs a warning to show that this is losing data |
|
pls do the tests like the others, e.g. use |
|
pls add a warning when things are stringified as well (as this is generally something which is not obvious and makes querying different). |
|
Do you think that it's better to just straight-up disallow |
|
I think it might be better to raise a more informative error message the point of the Table format is to query and so have integer names though can work is a bit odd |
|
@jameshiebert I think let's just raise an error message. Pls rebase and update. |
|
Sounds good. Closing this in favor of #10098 |
closes #9057
Converted column names to strings before creating pytables metadata object
This brute force
str()ing method might not be the best approach, but AFAIK it hasn't broken any of the tests.