DataFrame.convert_objects(copy=False) is ignored

[ankostis]

When invoking DataFrame.convert_objects() with copy=False, it does not affect affect the original object.

It should be:

1. either mentioned in the documentation (as Make convert_objects less aggressive #3436 suggest),
2. enforce it, or
3. drop the kw-arg.

    >>> import pandas as pd

    >>> df = pd.DataFrame([['1',2], [3,'4']])
    >>> df.dtypes
    Out[57]: 
    0    object
    1    object
    dtype: object
    >>> df.convert_objects(convert_numeric = True, copy = False)
    >>> df.dtypes
    Out[59]: 
    0    object
    1    object
    dtype: object

INSTALLED VERSIONS
------------------
commit: None
python: 3.3.3.final.0
python-bits: 64
OS: Windows
OS-release: 7
machine: AMD64
processor: Intel64 Family 6 Model 42 Stepping 7, GenuineIntel
byteorder: little
LC_ALL: None
LANG: None

pandas: 0.13.1
Cython: 0.20
numpy: 1.8.0
scipy: 0.13.3
statsmodels: 0.5.0
IPython: 1.1.0
sphinx: 1.2.1
patsy: 0.2.1
scikits.timeseries: None
dateutil: 2.2
pytz: 2013.9
bottleneck: None
tables: 3.1.0
numexpr: 2.3
matplotlib: 1.3.1
openpyxl: None
xlrd: 0.9.2
xlwt: None
xlsxwriter: None
sqlalchemy: 0.9.2
lxml: None
bs4: None
html5lib: None
bq: None
apiclient: None

[jreback]

its in the doc-string. Its not supposed to affect the original object, rather to simply return the original if no copy is necessary (e.g. no conversion was done). But honestly this method should not be used very often and conversion is done for you almost always. The copy kw is really for internal usage.

Can you give a context where you actually need this?

would certainly take a PR to expand doc-string, or give a better explanation in the docs

[ankostis]

Maybe you are talking about a different version.
I cannot see any such explanation you suggested in the docstring:

    >>> pd.DataFrame.convert_objects?
    Type:       function
    String Form:<function NDFrame.convert_objects at 0x0000000006BD7488>
    File:       d:\apps\winpython-64bit-3.3.3.3\python-3.3.3.amd64\lib\site-packages\pandas\core\generic.py
    Definition: pd.DataFrame.convert_objects(self, convert_dates=True, convert_numeric=False, convert_timedeltas=True, copy=True)
    Docstring:
    Attempt to infer better dtype for object columns

    Parameters
    ----------
    ....
    copy : Boolean, if True, return copy, default is True

In any case, i believe it needs a better doc-string, or/an a rename even better.

[jreback]

http://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.convert_objects.html?highlight=convert_objects#pandas.DataFrame.convert_objects

you must be looking at an older version

In [6]: pd.DataFrame.convert_objects?
Type:        instancemethod
String form: <unbound method DataFrame.convert_objects>
File:        /usr/local/lib/python2.7/site-packages/pandas-0.13.1-py2.7-linux-x86_64.egg/pandas/core/generic.py
Definition:  pd.DataFrame.convert_objects(self, convert_dates=True, convert_numeric=False, convert_timedeltas=True, copy=True)
Docstring:
Attempt to infer better dtype for object columns

Parameters
----------
convert_dates : if True, attempt to soft convert dates, if 'coerce',
    force conversion (and non-convertibles get NaT)
convert_numeric : if True attempt to coerce to numbers (including
    strings), non-convertibles get NaN
convert_timedeltas : if True, attempt to soft convert timedeltas, if 'coerce',
    force conversion (and non-convertibles get NaT)
copy : Boolean, if True, return copy, default is True

Returns
-------
converted : asm as input object

[jreback]

@ankostis I believe you are confusing copy with inplace. This always returns an object, never modifying the existing (which in general is not a good-practice in any event).

copy just determines if an 'extra' copy is made if it was forced to copy.

If you would like to submit a PR to modify the doc-string feel free.

[ankostis]

If you would like to submit a PR to modify the doc-string feel free.

done: PR #6791.

By the way, i was looking at the exact same version as you.

Thanks for you time.

[jreback]

was closed by #6791