Panel.describe() docs for v15.2 here

Environment:

fedora 20 pandas 15.2 Python 2.7.5 IPython 2.3.0

To duplicate:

$ pip install --upgrade pandas

In [8]: thresh = pd.Panel({'1': pd.DataFrame(pd.np.arange(2,3)), '2': pd.DataFrame(pd.np.arange(1,2))})

In [9]: thresh
Out[9]: 
<class 'pandas.core.panel.Panel'>
Dimensions: 2 (items) x 1 (major_axis) x 1 (minor_axis)
Items axis: 1 to 2
Major_axis axis: 0 to 0
Minor_axis axis: 0 to 0

In [10]: thresh['1'].describe()
Out[10]: 
        0
count   1
mean    2
std   NaN
min     2
25%     2
50%     2
75%     2
max     2

In [11]: thresh.describe()
---------------------------------------------------------------------------
NotImplementedError                       Traceback (most recent call last)
/usr/src/projects/predict/predict/webapp/dashboard/models.pyc in <module>()
----> 1 thresh.describe()

/usr/local/share/.virtualenvs/predict/lib/python2.7/site-packages/pandas/core/generic.pyc in describe(self, percentile_width, percentiles, include, exclude)
   3748         if self.ndim >= 3:
   3749             msg = "describe is not implemented on on Panel or PanelND objects."
-> 3750             raise NotImplementedError(msg)
   3751 
   3752         if percentile_width is not None and percentiles is not None:

NotImplementedError: describe is not implemented on on Panel or PanelND objects.

In [12]: pd.__version__
Out[12]: '0.15.2'

In [13]: type(thresh)
Out[13]: pandas.core.panel.Panel

Comment From: jreback

@hobson this actually is NotImplemented and that's why the error shows.

You could this.

In [19]: thresh.apply(lambda x: x.describe(),axis=['items','major'])[0]    
Out[19]: 
        1   2
count   1   1
mean    2   1
std   NaN NaN
min     2   1
25%     2   1
50%     2   1
75%     2   1
max     2   1

Comment From: hobson

Thank you for that implementation/work-around. If you dont mind, I'll issue a pr using your 'apply' code, so the code catches up with the docs. On Dec 16, 2014 3:17 AM, "jreback" notifications@github.com wrote:

@hobson https://github.com/hobson this actually is NotImplemented and that's why the error shows.

You could this.

In [19]: thresh.apply(lambda x: x.describe(),axis=['items','major'])[0] Out[19]: 1 2 count 1 1 mean 2 1 std NaN NaN min 2 1 25% 2 1 50% 2 1 75% 2 1 max 2 1

— Reply to this email directly or view it on GitHub https://github.com/pydata/pandas/issues/9091#issuecomment-67144883.

Comment From: jreback

closing as Panels are deprecated