Python: How to filter a DataFrame of dates in Pandas by a particular date within a window of some days?

Question

I have a DataFrame of dates and would like to filter for a particular date +- some days.

import pandas as pd
import numpy as np
import datetime

dates = pd.date_range(start="08/01/2009",end="08/01/2012",freq="D")
df = pd.DataFrame(np.random.rand(len(dates), 1)*1500, index=dates, columns=['Power'])

If I select lets say date 2009-08-03 and a window of 5 days, output would be similar to:

>>> 
                  Power
2010-07-29   713.108020
2010-07-30  1055.109543
2010-07-31   951.159099
2010-08-01  1350.638983
2010-08-02   453.166697
2010-08-03  1066.859386
2010-08-04  1381.900717
2010-08-05   107.489179
2010-08-06  1195.945723
2010-08-07  1209.762910
2010-08-08   349.554492

NB: The original problem I am trying to accomplish is under Python: Filter DataFrame in Pandas by hour, day and month grouped by year

Answer 1

The function I created to accomplish this is filterDaysWindow and can be used as follows:

import pandas as pd
import numpy as np
import datetime

dates = pd.date_range(start="08/01/2009",end="08/01/2012",freq="D")
df = pd.DataFrame(np.random.rand(len(dates), 1)*1500, index=dates, columns=['Power'])

def filterDaysWindow(df, date, daysWindow):
    """
    Filter a Dataframe by a date within a window of days

    @type df: DataFrame
    @param df: DataFrame of dates

    @type date: datetime.date
    @param date: date to focus on

    @type daysWindow: int
    @param daysWindow: Number of days to perform the days window selection

    @rtype: DataFrame
    @return: Returns a DataFrame with dates within date+-daysWindow
    """    
    dateStart = date - datetime.timedelta(days=daysWindow)
    dateEnd = date + datetime.timedelta(days=daysWindow)
    return df [dateStart:dateEnd]

df_filtered = filterDaysWindow(df, datetime.date(2010,8,3), 5)
print df_filtered