摘要:
六、日期時間預處理
作者:Chris Albon
譯者:飛龍
協議:CC BY-NC-SA 4.0
把日期和時間拆成多個特徵
# 載...
六、日期時間預處理
作者:ofollow,noindex">Chris Albon
譯者:飛龍
協議:CC BY-NC-SA 4.0
把日期和時間拆成多個特徵
# 載入庫
import pandas as pd
# 建立資料幀
df = pd.DataFrame()
# 建立五個日期
df['date'] = pd.date_range('1/1/2001', periods=150, freq='W')
# 為年月日,時分秒建立特徵
df['year'] = df['date'].dt.year
df['month'] = df['date'].dt.month
df['day'] = df['date'].dt.day
df['hour'] = df['date'].dt.hour
df['minute'] = df['date'].dt.minute
# 展示三行
df.head(3)
|
date |
year |
month |
day |
hour |
minute |
0 |
2001-01-07 |
2001 |
1 |
7 |
0 |
0 |
1 |
2001-01-14 |
2001 |
1 |
14 |
0 |
0 |
2 |
2001-01-21 |
2001 |
1 |
21 |
0 |
0 |
計算日期時間之間的差
# 載入庫
import pandas as pd
# 建立資料幀
df = pd.DataFrame()
# 建立兩個 datetime 特徵
df['Arrived'] = [pd.Timestamp('01-01-2017'), pd.Timestamp('01-04-2017')]
df['Left'] = [pd.Timestamp('01-01-2017'), pd.Timestamp('01-06-2017')]
# 計算特徵之間的間隔
df['Left'] - df['Arrived']
'''
00 days
12 days
dtype: timedelta64[ns]
'''
# 計算特徵之間的間隔
pd.Series(delta.days for delta in (df['Left'] - df['Arrived']))
'''
00
12
dtype: int64
'''
將字串轉換為日期
# 載入庫
import numpy as np
import pandas as pd
# 建立字串
date_strings = np.array(['03-04-2005 11:35 PM',
'23-05-2010 12:01 AM',
'04-09-2009 09:09 PM'])
如果errors="coerce"
那麼任何問題都不會產生錯誤(預設行為),而是將導致錯誤的值設定為NaT
(即缺失值)。
程式碼 |
描述 |
示例 |
%Y
|
整年 |
2001
|
%m
|
零填充的月份 |
04
|
%d
|
零填充的日期 |
09
|
%I
|
零填充的小時(12 小時) |
02
|
%p
|
AM 或 PM |
AM
|
%M
|
零填充的分鐘 |
05
|
%S
|
零填充的秒鐘 |
09
|
# 轉換為 datetime
[pd.to_datetime(date, format="%d-%m-%Y %I:%M %p", errors="coerce") for date in date_strings]
'''
[Timestamp('2005-04-03 23:35:00'),
Timestamp('2010-05-23 00:01:00'),
Timestamp('2009-09-04 21:09:00')]
'''
轉換 pandas 列的時區
# 載入庫
import pandas as pd
from pytz import all_timezones
# 展示十個時區
all_timezones[0:10]
'''
['Africa/Abidjan',
'Africa/Accra',
'Africa/Addis_Ababa',
'Africa/Algiers',
'Africa/Asmara',
'Africa/Asmera',
'Africa/Bamako',
'Africa/Bangui',
'Africa/Banjul',
'Africa/Bissau']
'''
# 建立十個日期
dates = pd.Series(pd.date_range('2/2/2002', periods=10, freq='M'))
# 設定時區
dates_with_abidjan_time_zone = dates.dt.tz_localize('Africa/Abidjan')
# 檢視 pandas 序列
dates_with_abidjan_time_zone
'''
02002-02-28 00:00:00+00:00
12002-03-31 00:00:00+00:00
22002-04-30 00:00:00+00:00
32002-05-31 00:00:00+00:00
42002-06-30 00:00:00+00:00
52002-07-31 00:00:00+00:00
62002-08-31 00:00:00+00:00
72002-09-30 00:00:00+00:00
82002-10-31 00:00:00+00:00
92002-11-30 00:00:00+00:00
dtype: datetime64[ns, Africa/Abidjan]
'''
# 轉換時區
dates_with_london_time_zone = dates_with_abidjan_time_zone.dt.tz_convert('Europe/London')
# 檢視 pandas 序列
dates_with_london_time_zone
'''
02002-02-28 00:00:00+00:00
12002-03-31 00:00:00+00:00
22002-04-30 01:00:00+01:00
32002-05-31 01:00:00+01:00
42002-06-30 01:00:00+01:00
52002-07-31 01:00:00+01:00
62002-08-31 01:00:00+01:00
72002-09-30 01:00:00+01:00
82002-10-31 00:00:00+00:00
92002-11-30 00:00:00+00:00
dtype: datetime64[ns, Europe/London]
'''
編碼星期
# 載入庫
import pandas as pd
# 建立資料集
dates = pd.Series(pd.date_range('2/2/2002', periods=3, freq='M'))
# 檢視資料
dates
'''
02002-02-28
12002-03-31
22002-04-30
dtype: datetime64[ns]
'''
# 檢視星期
dates.dt.weekday_name
'''
0Thursday
1Sunday
2Tuesday
dtype: object
'''
處理時間序列中的缺失值
# 載入庫
import pandas as pd
import numpy as np
# 建立日期
time_index = pd.date_range('01/01/2010', periods=5, freq='M')
# 建立資料幀,設定索引
df = pd.DataFrame(index=time_index)
# 建立帶有一些缺失值的特徵
df['Sales'] = [1.0,2.0,np.nan,np.nan,5.0]
# 對缺失值執行插值
df.interpolate()
|
Sales |
2010-01-31 |
1.0 |
2010-02-28 |
2.0 |
2010-03-31 |
3.0 |
2010-04-30 |
4.0 |
2010-05-31 |
5.0 |
# 前向填充
df.ffill()
|
Sales |
2010-01-31 |
1.0 |
2010-02-28 |
2.0 |
2010-03-31 |
2.0 |
2010-04-30 |
2.0 |
2010-05-31 |
5.0 |
# 後向填充
df.bfill()
|
Sales |
2010-01-31 |
1.0 |
2010-02-28 |
2.0 |
2010-03-31 |
5.0 |
2010-04-30 |
5.0 |
2010-05-31 |
5.0 |
# 對缺失值執行插值
df.interpolate(limit=1, limit_direction='forward')
|
Sales |
2010-01-31 |
1.0 |
2010-02-28 |
2.0 |
2010-03-31 |
3.0 |
2010-04-30 |
NaN |
2010-05-31 |
5.0 |
處理時區
# 載入庫
import pandas as pd
from pytz import all_timezones
# 展示十個時區
all_timezones[0:10]
'''
['Africa/Abidjan',
'Africa/Accra',
'Africa/Addis_Ababa',
'Africa/Algiers',
'Africa/Asmara',
'Africa/Asmera',
'Africa/Bamako',
'Africa/Bangui',
'Africa/Banjul',
'Africa/Bissau']
'''
# 建立 datetime
pd.Timestamp('2017-05-01 06:00:00', tz='Europe/London')
# Timestamp('2017-05-01 06:00:00+0100', tz='Europe/London')
# 建立 datetime
date = pd.Timestamp('2017-05-01 06:00:00')
# 設定時區
date_in_london = date.tz_localize('Europe/London')
# 修改時區
date_in_london.tz_convert('Africa/Abidjan')
# Timestamp('2017-05-01 05:00:00+0000', tz='Africa/Abidjan')
平移時間特徵
# 載入庫
import pandas as pd
# 建立資料幀
df = pd.DataFrame()
# 建立資料
df['dates'] = pd.date_range('1/1/2001', periods=5, freq='D')
df['stock_price'] = [1.1,2.2,3.3,4.4,5.5]
# 將值平移一行
df['previous_days_stock_price'] = df['stock_price'].shift(1)
# 展示資料幀
df
|
dates |
stock_price |
previous_days_stock_price |
0 |
2001-01-01 |
1.1 |
NaN |
1 |
2001-01-02 |
2.2 |
1.1 |
2 |
2001-01-03 |
3.3 |
2.2 |
3 |
2001-01-04 |
4.4 |
3.3 |
4 |
2001-01-05 |
5.5 |
4.4 |
滑動時間視窗
# 載入庫
import pandas as pd
# 建立 datetime
time_index = pd.date_range('01/01/2010', periods=5, freq='M')
# 建立資料幀,設定索引
df = pd.DataFrame(index=time_index)
# 建立特徵
df['Stock_Price'] = [1,2,3,4,5]
# 計算滑動均值
df.rolling(window=2).mean()
|
Stock_Price |
2010-01-31 |
NaN |
2010-02-28 |
1.5 |
2010-03-31 |
2.5 |
2010-04-30 |
3.5 |
2010-05-31 |
4.5 |
# 識別滑動時間視窗中的最大值
df.rolling(window=2).max()
|
Stock_Price |
2010-01-31 |
NaN |
2010-02-28 |
2.0 |
2010-03-31 |
3.0 |
2010-04-30 |
4.0 |
2010-05-31 |
5.0 |
選擇日期時間範圍
# 載入庫
import pandas as pd
# 建立資料幀
df = pd.DataFrame()
# 建立 datetime
df['date'] = pd.date_range('1/1/2001', periods=100000, freq='H')
如果資料幀未按時間索引,請使用此方法。
# 選擇兩個日期時間之間的觀測
df[(df['date'] > '2002-1-1 01:00:00') & (df['date'] <= '2002-1-1 04:00:00')]
|
date |
8762 |
2002-01-01 02:00:00 |
8763 |
2002-01-01 03:00:00 |
8764 |
2002-01-01 04:00:00 |
如果資料幀按時間索引,請使用此方法。
# 設定索引
df = df.set_index(df['date'])
# 選擇兩個日期時間之間的觀測
df.loc['2002-1-1 01:00:00':'2002-1-1 04:00:00']
|
date |
date |
|
2002-01-01 01:00:00 |
2002-01-01 01:00:00 |
2002-01-01 02:00:00 |
2002-01-01 02:00:00 |
2002-01-01 03:00:00 |
2002-01-01 03:00:00 |
2002-01-01 04:00:00 |
2002-01-01 04:00:00 |