Pandas-时间序列（一）-数据类型：TimeSeries（时间序列）【以 DatatimeIndex/PeriodIndex 为 Index 的 Series】【“时间戳”与“时期”之间的转换】

- 日理万妓 2023-09-29 09:31 23阅读 0赞

Timestamp、Period作为index时区别不大

*  Timestamp表示一个时间戳，是一个时间截面；
 *  Period是一个时期，是一个时间段！！

## 一、以DatetimeIndex（元素为：Timestamp）为index ##

以DatetimeIndex为index的Series,为TimeSries(时间序列)

import numpy as np
    import pandas as pd
    
    # pd.DatetimeIndex()
    # 直接生成时间戳索引，支持str、datetime.datetime
    # 单个时间戳为Timestamp，多个时间戳为DatetimeIndex
    rng = pd.DatetimeIndex(['12/1/2017', '12/2/2017', '12/3/2017'])
    print("rng = {0}, type(rng) = {1}".format(rng, type(rng)))
    print("rng[0] = {0}, type(rng[0]) = {1}".format(rng[0], type(rng[0])))
    print('-' * 100)
    
    # TimeSries
    # 以DatetimeIndex为index的Series,为TimeSries(时间序列)
    st = pd.Series(np.random.rand(len(rng)), index=rng)
    print("st = \n{0} \ntype(st) = {1}".format(st, type(st)))
    print('-' * 50)
    print("st.index = ", st.index)

打印结果：

rng = DatetimeIndex(['2017-12-01', '2017-12-02', '2017-12-03'], dtype='datetime64[ns]', freq=None), type(rng) = <class 'pandas.core.indexes.datetimes.DatetimeIndex'>
    rng[0] = 2017-12-01 00:00:00, type(rng[0]) = <class 'pandas._libs.tslibs.timestamps.Timestamp'>
    ----------------------------------------------------------------------------------------------------
    st = 
    2017-12-01    0.310079
    2017-12-02    0.571608
    2017-12-03    0.557365
    dtype: float64 
    type(st) = <class 'pandas.core.series.Series'>
    --------------------------------------------------
    st.index =  DatetimeIndex(['2017-12-01', '2017-12-02', '2017-12-03'], dtype='datetime64[ns]', freq=None)
    
    Process finished with exit code 0

## 二、以PeriodIndex（元素为：Period）为index ##

import numpy as np
    import pandas as pd
    
    # pd.period_range()创建时期范围
    
    # 数据格式为PeriodIndex，单个数值为Period
    periods = pd.period_range(start='1/1/2011', end='1/1/2012', freq='M')
    print("periods = {0}, type(periods) = {1}".format(periods, type(periods)))
    print("-" * 50)
    print("periods[0] = {0}, type(periods[0]) = {1}".format(periods[0], type(periods[0])))
    print("-" * 200)
    
    
    # 时间序列
    # Period('2011', freq = 'A-DEC')可以看成多个时间期的时间段中的游标
    # Timestamp表示一个时间戳，是一个时间截面；Period是一个时期，是一个时间段！！但两者作为index时区别不大
    ts = pd.Series(np.random.rand(len(periods)), index=periods)
    print("ts = \n{0} \ntype(ts) = {1}".format(ts, type(ts)))
    print("-" * 50)
    print("ts.index = \n", ts.index)

打印结果：

periods = PeriodIndex(['2011-01', '2011-02', '2011-03', '2011-04', '2011-05', '2011-06',
                 '2011-07', '2011-08', '2011-09', '2011-10', '2011-11', '2011-12',
                 '2012-01'],
                dtype='period[M]', freq='M'), type(periods) = <class 'pandas.core.indexes.period.PeriodIndex'>
    --------------------------------------------------
    periods[0] = 2011-01, type(periods[0]) = <class 'pandas._libs.tslibs.period.Period'>
    --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
    ts = 
    2011-01    0.454952
    2011-02    0.854056
    2011-03    0.255570
    2011-04    0.066163
    2011-05    0.313975
    2011-06    0.682848
    2011-07    0.494482
    2011-08    0.737051
    2011-09    0.883112
    2011-10    0.137797
    2011-11    0.255970
    2011-12    0.802554
    2012-01    0.091610
    Freq: M, dtype: float64 
    type(ts) = <class 'pandas.core.series.Series'>
    --------------------------------------------------
    ts.index = 
     PeriodIndex(['2011-01', '2011-02', '2011-03', '2011-04', '2011-05', '2011-06', '2011-07', '2011-08', '2011-09', '2011-10', '2011-11', '2011-12', '2012-01'],
                dtype='period[M]', freq='M')
    
    Process finished with exit code 0

## 三、时间戳与时期之间的转换：pd.to\_period()、pd.to\_timestamp() ##

import numpy as np
    import pandas as pd
    
    # 时间戳与时期之间的转换：pd.to_period()、pd.to_timestamp()
    
    rng = pd.date_range('2017/1/1', periods=10, freq='M')
    prng = pd.period_range('2017', '2018', freq='M')
    
    print("rng = {0} \ntype(rng) = {1}".format(rng, type(rng)))
    print("-" * 100)
    print("prng = {0} \ntype(prng) = {1}".format(prng, type(prng)))
    print("-" * 200)
    
    # 每月最后一日，转化为每月
    ts1 = pd.Series(np.random.rand(len(rng)), index=rng)
    print("type(ts1.index)) = {}".format(type(ts1.index)))
    print("ts1.head() = \n", ts1.head())
    print("-" * 100)
    ts11 = ts1.to_period()
    print("ts11 = ts1.to_period()---->type(ts11.index)) = {}".format(type(ts11.index)))
    print("ts11 = ts1.to_period()---->ts11.head() = \n", ts11.head())
    print("-" * 200)
    
    # 每月，转化为每月第一天
    ts2 = pd.Series(np.random.rand(len(prng)), index=prng)
    print("type(ts2.index)) = {}".format(type(ts2.index)))
    print("ts2.head() = \n", ts2.head())
    print("-" * 100)
    ts22 = ts2.to_timestamp()
    print("ts22 = ts2.to_timestamp()---->type(ts22.index)) = {}".format(type(ts22.index)))
    print("ts22 = ts2.to_timestamp()---->ts22.head() = \n", ts22.head())
    print("-" * 200)

打印结果：

rng = DatetimeIndex(['2017-01-31', '2017-02-28', '2017-03-31', '2017-04-30',
                   '2017-05-31', '2017-06-30', '2017-07-31', '2017-08-31',
                   '2017-09-30', '2017-10-31'],
                  dtype='datetime64[ns]', freq='M') 
    type(rng) = <class 'pandas.core.indexes.datetimes.DatetimeIndex'>
    ----------------------------------------------------------------------------------------------------
    prng = PeriodIndex(['2017-01', '2017-02', '2017-03', '2017-04', '2017-05', '2017-06',
                 '2017-07', '2017-08', '2017-09', '2017-10', '2017-11', '2017-12',
                 '2018-01'],
                dtype='period[M]', freq='M') 
    type(prng) = <class 'pandas.core.indexes.period.PeriodIndex'>
    --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
    type(ts1.index)) = <class 'pandas.core.indexes.datetimes.DatetimeIndex'>
    ts1.head() = 
     2017-01-31    0.307686
    2017-02-28    0.752647
    2017-03-31    0.378614
    2017-04-30    0.519277
    2017-05-31    0.382742
    Freq: M, dtype: float64
    ----------------------------------------------------------------------------------------------------
    ts11 = ts1.to_period()---->type(ts11.index)) = <class 'pandas.core.indexes.period.PeriodIndex'>
    ts11 = ts1.to_period()---->ts11.head() = 
     2017-01    0.307686
    2017-02    0.752647
    2017-03    0.378614
    2017-04    0.519277
    2017-05    0.382742
    Freq: M, dtype: float64
    --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
    type(ts2.index)) = <class 'pandas.core.indexes.period.PeriodIndex'>
    ts2.head() = 
     2017-01    0.804532
    2017-02    0.820868
    2017-03    0.065087
    2017-04    0.016507
    2017-05    0.040606
    Freq: M, dtype: float64
    ----------------------------------------------------------------------------------------------------
    ts22 = ts2.to_timestamp()---->type(ts22.index)) = <class 'pandas.core.indexes.datetimes.DatetimeIndex'>
    ts22 = ts2.to_timestamp()---->ts22.head() = 
     2017-01-01    0.804532
    2017-02-01    0.820868
    2017-03-01    0.065087
    2017-04-01    0.016507
    2017-05-01    0.040606
    Freq: MS, dtype: float64
    --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
    
    Process finished with exit code 0