Unclear error produced by resampling with NaT in Index

[patricktokeeffe]

Similar to #4746, trying to resample with an index containing NaT produces rather ambiguous error messages of the form:

ValueError: Shape of passed values is (2, 5), indices imply (2, 4)

Something like cannot resample from index containing Not-a-Time (NaT) would be more helpful to users.

MWE

import pandas as pd
import numpy as np

times = pd.DatetimeIndex(['20160428T110000',
                          '20160428T110001',
                          '20160428T110002',
                          '20160428T110003',
                          '20160428T110004',
                          pd.NaT,
                          '20160428T110001',
                          '20160428T110002',
                          '20160428T110003',
                          '20160428T110004'])

df = pd.DataFrame(np.random.randn(10,2),
                  index=times,
                  columns=['A', 'B'])
df.resample('3s')

This example index contains duplicate values, but results are the same w/ or w/o them.

output of pd.show_versions()

INSTALLED VERSIONS
------------------
commit: None
python: 2.7.9.final.0
python-bits: 32
OS: Windows
OS-release: 7
machine: AMD64
processor: Intel64 Family 6 Model 23 Stepping 10, GenuineIntel
byteorder: little
LC_ALL: None
LANG: None

pandas: 0.15.2
nose: 1.3.4
Cython: 0.21.1
numpy: 1.8.2
scipy: 0.14.0
statsmodels: 0.6.1
IPython: 2.3.1
sphinx: 1.2.3
patsy: 0.3.0
dateutil: 2.3
pytz: 2014.10
bottleneck: None
tables: 3.1.1
numexpr: 2.4
matplotlib: 1.4.2
openpyxl: None
xlrd: 0.9.3
xlwt: None
xlsxwriter: 0.6.4
lxml: 3.4.1
bs4: 4.3.2
html5lib: None
httplib2: None
apiclient: None
rpy2: 2.5.2
sqlalchemy: 0.9.8
pymysql: None
psycopg2: None

[jreback]

so this works in master. Note the syntax change in 0.18.0 (you have a quite old version).

In [23]: df.resample('3s').mean()
Out[23]: 
                            A         B
2016-04-28 11:00:00  0.633764 -0.551855
2016-04-28 11:00:03  0.024339  0.041184

I don't know if we are specifically testing for this. Welcome to have some confirming tests.

[jreback]

In [24]: df = pd.DataFrame(np.repeat(np.arange(10),2).reshape(10,-1),columns=list('AB'),index=times)

In [25]: df
Out[25]: 
                     A  B
2016-04-28 11:00:00  0  0
2016-04-28 11:00:01  1  1
2016-04-28 11:00:02  2  2
2016-04-28 11:00:03  3  3
2016-04-28 11:00:04  4  4
NaT                  5  5
2016-04-28 11:00:01  6  6
2016-04-28 11:00:02  7  7
2016-04-28 11:00:03  8  8
2016-04-28 11:00:04  9  9

In [26]: df.resample('3s').mean()
Out[26]: 
                       A    B
2016-04-28 11:00:00  3.2  3.2
2016-04-28 11:00:03  6.0  6.0