Timestamp tiimestamp: fixed time -> pd.Timestamp
period fixed period of time: for example, in March 2016, another example, 2015 sales -> pd.Period
Interval interval: is represented by the start time and end time, a fixed period is a specific time interval
The date and time in the role of Pandas: analysis of financial data, such as stock transaction data
. 1 Import PANDAS AS PD 2 Import numpy AS NP . 3 . 4 # processing time need to use the package . 5 from datetime Import datetime . 6 from datetime Import timedelta . 7 . 8 now DateTime.Now = () # outputs the current time 9 # e.g. output datetime. datetime (2019,. 8, 23 is,. 17, 45, 13 is, 291738) 10 now.year, now.month, now.day # output date (2019,. 8, 23 is) . 11 12 is DATAl = datetime (2016,. 4, 20 is ) 13 is DATA2 = datetime (2016,. 4, 16 ) 14 Delta = DATAl - DATA2 # outputs the datetime.timedelta (. 4) 15 delta.days # output. 4 16 delta.total_seconds () # output interval 345,600.0 seconds . 17 DATA2 Delta + # Returns the day's date datetime.datetime (2016, 4, 20 is, 0, 0) 18 is DATA2 + timedelta (4.5 of 5) # 4.5 of 5 days, it outputs A datetime.datetime (2016,. 4, 20 is, 12 is, 0) . 19 20 is DATE = datetime (2016,3,20,8,30 ) 21 is STR (date) # converted into a string date '2016-03-20 08:30:00' 22 is date.strftime ( " % the Y /% m /% D% H:% m:% S " ) #Formatted output '2016/03/20 08:30:00' 23 is 24 datetime.strptime ( ' 2016-03-20 09:30 ' , ' % Y-M-% D%% H:% M ' ) 25 # deformatter output A datetime.datetime (2016,. 3, 20 is,. 9, 30) 26 is 27 a dates = [datetime (2016,3,1), datetime (2016,3,2 ), 28 datetime (2016,3,3) , datetime (2016,3,4 )] 29 S = pd.Series (np.random.randn (. 4), index = a dates) 30 # the time becomes the index to Series way output 31 is type (s.index) # it the return value pandas.core.indexes.datetimes.DatetimeIndex 32 type (s.index [0]) # The return value pandas._libs.tslibs.timestamps.Timestamp 33 is 34 is pd.date_range ( ' 20.16032 million ' , ' 20.16033 million ' ) # output time list, the default is the frequency interval / D 35 36 pd.date_range ( ' 20.16032 million 16:32 : 38 is ' , = 10 periods, = the normalize True) 37 [ # output frequency intervals of 10 days by default / D per day i.e., using regularization normalize = True removed 16:32:38 38 is 39 pd.date_range ( ' 20.16032 million ' , periods = 10, FREQ = ' M ' ) 40 # output end time of the month, a total of 10 months, monthly interval frequency is 41 #freq = BM (the last business day of each month), W weekly frequency, 4H 4 hours frequency 42 is 43 is 44 is 45 46 is P = pd.Period (2010, FREQ = ' M ' ) # in months period output Period ( '2010-01', 'M') 47 P + 2 # will lose output Period ( '2010-03', 'M') 48 49 pd.period_range ( ' 2016-01 ' , 10 periods = , FREQ = ' M ' ) 50 # at month intervals for 10 consecutive months frequency output 51 is 52 is pd.period_range ( ' 2016Q1 ' , = 10 periods, FREQ = ' Q ') 53 #Season for the frequency of the output at intervals of 10 consecutive seasons 54 is 55 56 is A = pd.Period (2016) # The default frequency band of 57 is a.asfreq ( ' M ' ) # Output Period ( '2016-12', 'M ') to the last one month period default band that becomes dated 58 a.asfreq ( ' M ' , How = ' Start ' ) # period becomes January period (' 2016-01 ',' M ') 59 P = PD .Period ( ' 2016-04 ' , FREQ = ' M ' ) # custom period period ( '2016-04', 'M') 60 61 is # 62 is P.asfreq ( 'DEC-A ' ) # frequency is converted to an annual basis output Period ending December (' 2016 ',' A-DEC ') 63 p.asfreq ( ' A-MAR ' ) # intervene month time becomes March to March 2017 64- 65 the p-pd.Period = ( ' 2016Q4 ' , ' Q-JAN ' ) # to the quarter ended January 66 # next line can know the start time and end time 2016Q4 67 p.asfreq ( ' M ' , How = ' Start ' ), p.asfreq ( ' M ' , How = ' End ') 68 69 # Get the penultimate quarter days 16:20 70 (p.asfreq ( ' B ' ) -. 1) .asfreq ( ' T ' ) + 16 * 60 +20