Date Arithmetic

documentation: http://pandas.pydata.org/pandas-docs/stable/timeseries.html#timeseries-offsets

Type		Description
date		Store calendar date (year, month, day) using a Gregorian Calendar
datetime		Store both date and time
timedelta		Difference between two datetime values

common date arithmetic operations

calculate differences between date
generate sequences of dates and time spans
convert time series to a particular frequency

Date, time, functions

documentation: http://pandas.pydata.org/pandas-docs/stable/api.html#top-level-dealing-with-datetimelike

to_datetime(args, *kwargs)	Convert argument to datetime.
to_timedelta(args, *kwargs)	Convert argument to timedelta
date_range([start, end, periods, freq, tz, ...])	Return a fixed frequency datetime index, with day (calendar) as the default
bdate_range([start, end, periods, freq, tz, ...])	Return a fixed frequency datetime index, with business day as the default
period_range([start, end, periods, freq, name])	Return a fixed frequency datetime index, with day (calendar) as the default
timedelta_range([start, end, periods, freq, ...])	Return a fixed frequency timedelta index, with day as the default
infer_freq(index[, warn])	Infer the most likely frequency given the input index.



In [3]:

    
import pandas as pd
import numpy as np
from datetime import datetime

now()



In [4]:

    
now = datetime.now()
now









    Out[4]:





datetime.datetime(2016, 8, 16, 14, 12, 8, 750310)



In [5]:

    
now.year, now.month, now.day









    Out[5]:





(2016, 8, 16)

delta

source: http://pandas.pydata.org/pandas-docs/stable/timedeltas.html



In [6]:

    
delta = now - datetime(2001, 1, 1)
delta









    Out[6]:





datetime.timedelta(5706, 51128, 750310)



In [7]:

    
delta.days









    Out[7]:





5706

Parsing Timedelta

from string



In [8]:

    
pd.Timedelta('4 days 7 hours')









    Out[8]:





Timedelta('4 days 07:00:00')

named keyword arguments



In [9]:

    
# note: these MUST be specified as keyword arguments
pd.Timedelta(days=1, seconds=1)









    Out[9]:





Timedelta('1 days 00:00:01')

integers with a unit



In [10]:

    
pd.Timedelta(1, unit='d')









    Out[10]:





Timedelta('1 days 00:00:00')

create a range of dates from Timedelta



In [11]:

    
us_memorial_day = datetime(2016, 5, 30)
print(us_memorial_day)
us_labor_day = datetime(2016, 9, 5)
print(us_labor_day)
us_summer_time = us_labor_day - us_memorial_day
print(us_summer_time)
type(us_summer_time)









    



2016-05-30 00:00:00
2016-09-05 00:00:00
98 days, 0:00:00






    Out[11]:





datetime.timedelta



In [12]:

    
us_summer_time_range = pd.date_range(us_memorial_day, periods=us_summer_time.days, freq='D')

summer_time time series with random data



In [13]:

    
us_summer_time_time_series = pd.Series(np.random.randn(us_summer_time.days), index=us_summer_time_range)
us_summer_time_time_series.tail()









    Out[13]:





2016-08-31   -0.347080
2016-09-01   -0.266705
2016-09-02    0.337686
2016-09-03   -0.210966
2016-09-04   -1.593387
Freq: D, dtype: float64



In [ ]: