CODEWITHSUNDEEP
pythonpandasnumpy
Kapil Sharma
Kapil Sharma

Reputation: 1432

Exclude between time in pandas

I know that you can select data from pandas.DatetimeIndex using pandas.DataFrame.between_time. Is there a convenient way to exclude between two times in pandas?

For example, to exclude data between 16:00 and 17:00, I am currently doing the following.

In [1]: import pandas as pd
        import numpy as np

In [2]: df = pd.DataFrame(np.random.randn(24 * 60 + 1, 2), columns=list("AB"), index=pd.date_range(start="20161013 00:00:00", freq="1T", periods=24 * 60 +1))

In [3]: idx = df.index.hour == 16

In [4]: df = df[~idx]

In [5]: df.between_time("16:00", "17:00")
Out[5]: 
                            A         B
2016-10-13 17:00:00 -0.745892  1.832912

EDIT

I have been able to use this:

In[41]:df2 = df.ix[np.setdiff1d(df.index, df.between_time("16:00", "17:00").index)]


In[42]:df2.between_time("15:59", "17:01")
Out[42]: 
                            A         B
2016-10-13 15:59:00  1.190678  0.783776
2016-10-13 17:01:00 -0.590931 -1.059962

Is there a better way?

Upvotes: 6

Views: 4786

Answers (2)

root
root

Reputation: 33783

You can combine between_time with drop:

df2 = df.drop(df.between_time("16:00", "17:00").index)

Edit

An alternate method is to exploit the fact that between_time operates circularly, so you can switch the order of your input times to exclude the range between them:

df.between_time("17:00", "16:00", include_start=False, include_end=False)

Upvotes: 13

David
David

Reputation: 11573

df['hour'] = df.index.hour
df[(df['hour'] < 16) | (df['hour'] > 17)]

Upvotes: 0

Related Questions