"df.get(None)" throws error in 0.13 -> regression?

[jankatins]

This works with pandas 0.12:

from ggplot import meat
meat.get(None)
[returns None]

But throws an error in 0.13 (RC):

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-11-86be6a7317b8> in <module>()
----> 1 meat.get(None)

C:\portabel\Python27\lib\site-packages\pandas\core\generic.pyc in get(self, key, default)
    961         """
    962         try:
--> 963             return self[key]
    964         except KeyError:
    965             return default

C:\portabel\Python27\lib\site-packages\pandas\core\frame.pyc in __getitem__(self, key)
   1626             return self._getitem_multilevel(key)
   1627         else:
-> 1628             return self._getitem_column(key)
   1629 
   1630     def _getitem_column(self, key):

C:\portabel\Python27\lib\site-packages\pandas\core\frame.pyc in _getitem_column(self, key)
   1633         # get column
   1634         if self.columns.is_unique:
-> 1635             return self._get_item_cache(key)
   1636 
   1637         # duplicate columns & possible reduce dimensionaility

C:\portabel\Python27\lib\site-packages\pandas\core\generic.pyc in _get_item_cache(self, item)
    972         res = cache.get(item)
    973         if res is None:
--> 974             values = self._data.get(item)
    975             res = self._box_item_values(item, values)
    976             cache[item] = res

C:\portabel\Python27\lib\site-packages\pandas\core\internals.pyc in get(self, item)
   2737             if isnull(item):
   2738                 indexer = np.arange(len(self.items))[isnull(self.items)]
-> 2739                 return self.get_for_nan_indexer(indexer)
   2740 
   2741             _, block = self._find_block(item)

C:\portabel\Python27\lib\site-packages\pandas\core\internals.pyc in get_for_nan_indexer(self, indexer)
   2789                 indexer = indexer.item()
   2790             else:
-> 2791                 raise ValueError("cannot label index with a null key")
   2792 
   2793         # take a nan indexer and return the values

ValueError: cannot label index with a null key

Not sure if that is intentional as this broke some code in the facet_wrap in ggplot (which now will get some update against this :-) )

[jreback]

that seems like a very odd thing to do as its trying to get something indexed by None

@jreback , close as wontfix? this looks like reasonable behavior to me

[jreback]

what is meat? a sub-class of frame?

[jreback]

?

In [62]: DataFrame().get(None)
ValueError: cannot label index with a null key

[jreback]

easy to fix I think

https://twitter.com/johnmyleswhite/status/420247553922183170
context.

[jankatins]

meat is a dataframe which is packaged with ggplot. Thanks for the fix!

[jreback]

grt...pls test if you can (I found that problem, but doens't hurt to do this)

[bnguyenvanyen]

Hi, the behaviour of df.get(None) seems inconsistent.
The previous compatibility fix caught ValueError: cannot label index with a null key.
On my dataframe however (too many columns ?), what gets raised is
TypeError: cannot label index with a null key, at line 3605 of core/internals.py

Could this get caught as well or does it break something ?

Thank you

[jorisvandenbossche]

@bennguvaye if you think there is still a bug for a certain case, please open a new issue and add a small reproducible code example.