Using pandas isin with selecting dates

Question

A similar question like mine has been asked before, but unfortunately the answer was unable to help my problem. I have two sets of data columns: 1. the first is generated as follows:

t = pd.bdate_range(start='2016-05-16', end='2016-06-15')

This gives:

DatetimeIndex: DatetimeIndex(['2016-05-16', '2016-05-17', '2016-05-18', ...

The second is selected from an array:

v = myArray['Date']

This gives:

Series: 0 2016-05-17
1 2016-05-17
2 2016-05-17

This second output of data is a large series of about 40k entries. The reason why the dates don't change in this example is because there are epochs every two minutes therefore the date is the same until midnight. Since this array is currently from 2016-05-17 to 2016-06-17, I'd like to reduce this so that business days are only considered. I've tried:

h = pd.DataFrame(t)
u = pd.Series(t)
k = pd.DataFrame(v)
w = pd.Series(v)

Then I've tried various permutations of the isin method:

w.isin(t)
w.isin(t.values)
k.isin(t.values)

I'm just not seeing any booleans at the beginning of the isin check that are True. All seem to be false. Where am I going wrong? In order to get this to work, do I need to take much more care of the format of the arrays?

Please post sample data set for `myArray['Date']` and desired output. [How to make good reproducible pandas examples](http://stackoverflow.com/questions/20109391/how-to-make-good-reproducible-pandas-examples) — MaxU - stand with Ukraine, Jun 17 '16 at 11:31
I already did, it's above after I wrote "This gives:" Series: 0 2016-05-17..." — pymat, Jun 17 '16 at 12:51

score 0 · Answer 1 · answered Jun 17 '16 at 08:30

You are looking for v.weekday < 5.

v = pd.date_range(start='2016-05-16', end='2016-06-15', freq='2H')
v[v.weekday < 5]

DatetimeIndex(['2016-05-16 00:00:00', '2016-05-16 02:00:00',
               '2016-05-16 04:00:00', '2016-05-16 06:00:00',
               '2016-05-16 08:00:00', '2016-05-16 10:00:00',
               '2016-05-16 12:00:00', '2016-05-16 14:00:00',
               '2016-05-16 16:00:00', '2016-05-16 18:00:00',
               ...
               '2016-06-14 06:00:00', '2016-06-14 08:00:00',
               '2016-06-14 10:00:00', '2016-06-14 12:00:00',
               '2016-06-14 14:00:00', '2016-06-14 16:00:00',
               '2016-06-14 18:00:00', '2016-06-14 20:00:00',
               '2016-06-14 22:00:00', '2016-06-15 00:00:00'],
              dtype='datetime64[ns]', length=265, freq=None)

It's not obvious to see, but these are only business days.

Using pandas isin with selecting dates

1 Answers1