Interpreting passed slicers in using .loc(axis=1) |#AssertionError: Start slice bound is non-scalar

[dickster77]

From SO

The documentation suggests:

You can also specify the axis argument to .loc to interpret the passed slicers on a single axis.
see http://pandas.pydata.org/pandas-docs/stable/advanced.html#using-slicers

However I get an error trying to slice along the column index.

    import pandas as pd
    import numpy as np

    cols= [(yr,m) for yr in [2014,2015] for m in [7,8,9,10]]
    df = pd.DataFrame(np.random.randint(1,100,(10,8)),index=tuple('ABCDEFGHIJ'))
    df.columns =pd.MultiIndex.from_tuples(cols)

    print df.head()

      2014             2015            
        7   8   9   10   7   8   9   10
    A   68  51   6  48   24   3   4  85
    B   79  75  68  62   19  40  63  45
    C   60  15  32  32   37  95  56  38
    D    4  54  81  50   13  64  65  13
    E   78  21  84   1   83  18  39  57


    #This does not work as expected
    print df.loc(axis=1)[(2014,9):(2015,8)]

    #AssertionError: Start slice bound is non-scalar

    #but an arbitrary transpose and changing axis works!
    df = df.T
    print df.loc(axis=0)[(2014,9):(2015,8)]

              A   B   C   D   E   F   G   H   I   J
    2014 9    6  68  32  81  84  60  83  39  94  93
         10  48  62  32  50   1  84  18  14  92  33
    2015 7   24  19  37  13  83  69  31  91  69  90
         8    3  40  95  64  18   8  32  93  16  25

    #So I could always assign the slice and re-transpose. 
    #That though feels like a hack and the axis=1 setting should have worked.

        df = df.loc(axis=0)[(2014,9):(2015,8)]

    df = df.T


    print df

      2014     2015    
        9   10   7   8 
    A   64  98   99  87
    B   43  36   22  84
    C   32  78   86  66
    D   67   8   34  73
    E   83  54   96  33
    F   18  83   36  71
    G   13  25   76   8
    H   69   4   99  84
    I    3  52   50  62
    J   67  60    9  49

The canonical way to do this is:

In [6]: df.loc()[:,(2014,9):(2015,8)]
Out[6]: 
  2014     2015    
    9   10   7   8 
A   26   2   44  69
B   41   7    5   1
C    8  27   23  22
D   54  72   81  93
E   18  23   54   7
F   11  81   37  83
G   60  38   59  29
H    3  95   89  96
I    6   9   77   9
J   90  92   10  32

So prob a bug

[mroeschke]

This works on master. Could use a test

In [88]: df.loc(axis=1)[(2014,9):(2015,8)]
Out[88]:
  2014     2015
    9   10   7   8
A   65  14   26  75
B   39  69   54  49
C   90  73   11   9
D   32  31   15  14
E   52  88   46  14
F   93  91    1  28
G   36  69   43  88
H   15  68   38  12
I   53  30   20  26
J   51   5    6  96

In [82]: pd.__version__
Out[82]: '0.26.0.dev0+490.g9cfb8b55b'