只使用适用于DatetimeIndex的方法

# 读取crime hdf5数据集，行索引设为REPORTED_DATE，检查其数据类型
crime = pd.read_hdf('data/crime.h5', 'crime').set_index('REPORTED_DATE')
print(type(crime.index))
#<class 'pandas.core.indexes.datetimes.DatetimeIndex'>

# 用between_time方法选取发生在凌晨2点到5点的案件
crime.between_time('2:00', '5:00', include_end=False).head()

	OFFENSE_TYPE_ID	OFFENSE_CATEGORY_ID	GEO_LON	GEO_LAT	NEIGHBORHOOD_ID	IS_CRIME	IS_TRAFFIC
REPORTED_DATE
2014-06-29 02:01:00	traffic-accident-dui-duid	traffic-accident	-105.000149	39.745753	cbd	0	1
2014-06-29 02:00:00	disturbing-the-peace	public-disorder	-105.020719	39.706674	athmar-park	1	0
2014-06-29 02:18:00	curfew	public-disorder	-105.001552	39.769505	sunnyside	1	0
2014-06-29 04:17:00	aggravated-assault	aggravated-assault	-105.018557	39.679229	college-view-south-platte	1	0
2014-06-29 04:22:00	violation-of-restraining-order	all-other-crimes	-104.972447	39.739449	cheesman-park	1	0

# 用at_time方法选取特定时间
crime.at_time('5:47').head()

	OFFENSE_TYPE_ID	OFFENSE_CATEGORY_ID	GEO_LON	GEO_LAT	NEIGHBORHOOD_ID	IS_CRIME	IS_TRAFFIC
REPORTED_DATE
2013-11-26 05:47:00	criminal-mischief-other	public-disorder	-104.991476	39.751536	cbd	1	0
2017-04-09 05:47:00	criminal-mischief-mtr-veh	public-disorder	-104.959394	39.678425	university	1	0
2017-02-19 05:47:00	criminal-mischief-other	public-disorder	-104.986767	39.741336	north-capitol-hill	1	0
2017-02-16 05:47:00	aggravated-assault	aggravated-assault	-104.934029	39.732320	hale	1	0
2017-02-12 05:47:00	police-interference	all-other-crimes	-104.976306	39.722644	speer	1	0

# first方法可以选取排在前面的n个时间
# 首先将时间索引排序，然后使用pd.offsets模块
crime_sort = crime.sort_index()
pd.options.display.max_rows = 6
crime_sort.first(pd.offsets.MonthBegin(6))
#offset 是针对datetime格式的index进行范围限定，如果dataFrame的序列未按照时间排序，
#它也会筛选出来，但是通过int，窗口就是顺序计数。

	OFFENSE_TYPE_ID	OFFENSE_CATEGORY_ID	GEO_LON	GEO_LAT	NEIGHBORHOOD_ID	IS_CRIME	IS_TRAFFIC
REPORTED_DATE
2012-01-02 00:06:00	aggravated-assault	aggravated-assault	-104.816860	39.796717	montbello	1	0
2012-01-02 00:06:00	violation-of-restraining-order	all-other-crimes	-104.816860	39.796717	montbello	1	0
2012-01-02 00:16:00	traffic-accident-dui-duid	traffic-accident	-104.971851	39.736874	cheesman-park	0	1
...	...	...	...	...	...	...	...
2012-06-30 23:50:00	criminal-mischief-mtr-veh	public-disorder	-104.838271	39.788683	montbello	1	0
2012-06-30 23:54:00	traffic-accident-hit-and-run	traffic-accident	-105.014162	39.740439	lincoln-park	0	1
2012-07-01 00:01:00	robbery-street	robbery	-104.924292	39.767585	northeast-park-hill	1	0

27489 rows × 7 columns

# 前面的结果最后一条是7月的数据，这是因为pandas使用的是行索引中的第一个值，也就是2012-01-02 00:06:00
# 下面使用MonthEnd
crime_sort.first(pd.offsets.MonthEnd(6))

	OFFENSE_TYPE_ID	OFFENSE_CATEGORY_ID	GEO_LON	GEO_LAT	NEIGHBORHOOD_ID	IS_CRIME	IS_TRAFFIC
REPORTED_DATE
2012-01-02 00:06:00	aggravated-assault	aggravated-assault	-104.816860	39.796717	montbello	1	0
2012-01-02 00:06:00	violation-of-restraining-order	all-other-crimes	-104.816860	39.796717	montbello	1	0
2012-01-02 00:16:00	traffic-accident-dui-duid	traffic-accident	-104.971851	39.736874	cheesman-park	0	1
...	...	...	...	...	...	...	...
2012-06-29 23:41:00	robbery-street	robbery	-104.991912	39.756163	five-points	1	0
2012-06-29 23:57:00	assault-simple	other-crimes-against-persons	-104.987360	39.715162	speer	1	0
2012-06-30 00:04:00	traffic-accident	traffic-accident	-104.894697	39.628902	hampden-south	0	1

27332 rows × 7 columns

# 上面的结果中，6月30日的数据只有一条，这也是因为第一个时间值的原因。
# 所有的DateOffsets对象都有一个normalize参数，当其设为True时，会将所有时间归零。
# 下面就是我们想获得的结果
crime_sort.first(pd.offsets.MonthBegin(6, normalize=True))

OFFENSE_TYPE_ID	OFFENSE_CATEGORY_ID	GEO_LON	GEO_LAT	NEIGHBORHOOD_ID	IS_CRIME	IS_TRAFFIC
REPORTED_DATE
2012-01-02 00:06:00	aggravated-assault	aggravated-assault	-104.816860	39.796717	montbello	1	0
2012-01-02 00:06:00	violation-of-restraining-order	all-other-crimes	-104.816860	39.796717	montbello	1	0
2012-01-02 00:16:00	traffic-accident-dui-duid	traffic-accident	-104.971851	39.736874	cheesman-park	0	1
...	...	...	...	...	...	...	...
2012-06-30 23:44:00	traffic-accident	traffic-accident	-104.987578	39.711158	baker	0	1
2012-06-30 23:50:00	criminal-mischief-mtr-veh	public-disorder	-104.838271	39.788683	montbello	1	0
2012-06-30 23:54:00	traffic-accident-hit-and-run	traffic-accident	-105.014162	39.740439	lincoln-park	0	1

27488 rows × 7 columns

# 选取2012年1月到6月的数据
crime_sort.loc[:'2012-06']

	OFFENSE_TYPE_ID	OFFENSE_CATEGORY_ID	GEO_LON	GEO_LAT	NEIGHBORHOOD_ID	IS_CRIME	IS_TRAFFIC
REPORTED_DATE
2012-01-02 00:06:00	aggravated-assault	aggravated-assault	-104.816860	39.796717	montbello	1	0
2012-01-02 00:06:00	violation-of-restraining-order	all-other-crimes	-104.816860	39.796717	montbello	1	0
2012-01-02 00:16:00	traffic-accident-dui-duid	traffic-accident	-104.971851	39.736874	cheesman-park	0	1
...	...	...	...	...	...	...	...
2012-06-30 23:44:00	traffic-accident	traffic-accident	-104.987578	39.711158	baker	0	1
2012-06-30 23:50:00	criminal-mischief-mtr-veh	public-disorder	-104.838271	39.788683	montbello	1	0
2012-06-30 23:54:00	traffic-accident-hit-and-run	traffic-accident	-105.014162	39.740439	lincoln-park	0	1

27488 rows × 7 columns

## 5天
crime_sort.first('5D')

	OFFENSE_TYPE_ID	OFFENSE_CATEGORY_ID	GEO_LON	GEO_LAT	NEIGHBORHOOD_ID	IS_CRIME	IS_TRAFFIC
REPORTED_DATE
2012-01-02 00:06:00	aggravated-assault	aggravated-assault	-104.816860	39.796717	montbello	1	0
2012-01-02 00:06:00	violation-of-restraining-order	all-other-crimes	-104.816860	39.796717	montbello	1	0
2012-01-02 00:16:00	traffic-accident-dui-duid	traffic-accident	-104.971851	39.736874	cheesman-park	0	1
...	...	...	...	...	...	...	...
2012-01-06 23:30:00	assault-dv	other-crimes-against-persons	-104.958983	39.674135	university-park	1	0
2012-01-06 23:44:00	theft-of-motor-vehicle	auto-theft	-104.845356	39.794035	montbello	1	0
2012-01-06 23:55:00	threats-to-injure	public-disorder	-105.004788	39.708714	athmar-park	1	0

605 rows × 7 columns

## 5个工作日
crime_sort.first('5B')

	OFFENSE_TYPE_ID	OFFENSE_CATEGORY_ID	GEO_LON	GEO_LAT	NEIGHBORHOOD_ID	IS_CRIME	IS_TRAFFIC
REPORTED_DATE
2012-01-02 00:06:00	aggravated-assault	aggravated-assault	-104.816860	39.796717	montbello	1	0
2012-01-02 00:06:00	violation-of-restraining-order	all-other-crimes	-104.816860	39.796717	montbello	1	0
2012-01-02 00:16:00	traffic-accident-dui-duid	traffic-accident	-104.971851	39.736874	cheesman-park	0	1
...	...	...	...	...	...	...	...
2012-01-08 23:52:00	theft-other	larceny	-104.968227	39.739752	cheesman-park	1	0
2012-01-09 00:04:00	traffic-accident-hit-and-run	traffic-accident	-104.973343	39.760757	five-points	0	1
2012-01-09 00:05:00	fraud-criminal-impersonation	white-collar-crime	-105.024676	39.712702	valverde	1	0

879 rows × 7 columns

## 7周
crime_sort.first('7W')

	OFFENSE_TYPE_ID	OFFENSE_CATEGORY_ID	GEO_LON	GEO_LAT	NEIGHBORHOOD_ID	IS_CRIME	IS_TRAFFIC
REPORTED_DATE
2012-01-02 00:06:00	aggravated-assault	aggravated-assault	-104.816860	39.796717	montbello	1	0
2012-01-02 00:06:00	violation-of-restraining-order	all-other-crimes	-104.816860	39.796717	montbello	1	0
2012-01-02 00:16:00	traffic-accident-dui-duid	traffic-accident	-104.971851	39.736874	cheesman-park	0	1
...	...	...	...	...	...	...	...
2012-02-18 22:20:00	traffic-accident-dui-duid	traffic-accident	-104.919946	39.761917	north-park-hill	0	1
2012-02-18 22:44:00	criminal-mischief-mtr-veh	public-disorder	-105.044984	39.736776	west-colfax	1	0
2012-02-18 23:27:00	theft-items-from-vehicle	theft-from-motor-vehicle	-105.009018	39.708701	athmar-park	1	0

6708 rows × 7 columns

# 第3季度开始
crime_sort.first('3QS') #Quarterly 季度

	OFFENSE_TYPE_ID	OFFENSE_CATEGORY_ID	GEO_LON	GEO_LAT	NEIGHBORHOOD_ID	IS_CRIME	IS_TRAFFIC
REPORTED_DATE
2012-01-02 00:06:00	aggravated-assault	aggravated-assault	-104.816860	39.796717	montbello	1	0
2012-01-02 00:06:00	violation-of-restraining-order	all-other-crimes	-104.816860	39.796717	montbello	1	0
2012-01-02 00:16:00	traffic-accident-dui-duid	traffic-accident	-104.971851	39.736874	cheesman-park	0	1
...	...	...	...	...	...	...	...
2012-09-30 23:29:00	theft-of-motor-vehicle	auto-theft	-104.988838	39.686925	overland	1	0
2012-09-30 23:41:00	traffic-accident-hit-and-run	traffic-accident	-105.087598	39.638462	marston	0	1
2012-09-30 23:43:00	robbery-business	robbery	-104.772712	39.781966	gateway-green-valley-ranch	1	0

43045 rows × 7 columns

原理

# 使用datetime模块的time对象
import datetime
crime.between_time(datetime.time(2,0), datetime.time(5,0), include_end=False)

	OFFENSE_TYPE_ID	OFFENSE_CATEGORY_ID	GEO_LON	GEO_LAT	NEIGHBORHOOD_ID	IS_CRIME	IS_TRAFFIC
REPORTED_DATE
2014-06-29 02:01:00	traffic-accident-dui-duid	traffic-accident	-105.000149	39.745753	cbd	0	1
2014-06-29 02:00:00	disturbing-the-peace	public-disorder	-105.020719	39.706674	athmar-park	1	0
2014-06-29 02:18:00	curfew	public-disorder	-105.001552	39.769505	sunnyside	1	0
...	...	...	...	...	...	...	...
2017-09-13 02:21:00	assault-simple	other-crimes-against-persons	-104.925733	39.654184	university-hills	1	0
2017-09-13 03:21:00	traffic-accident-dui-duid	traffic-accident	-105.010711	39.757385	highland	0	1
2017-09-13 02:15:00	traffic-accident-hit-and-run	traffic-accident	-105.043950	39.787436	regis	0	1

29078 rows × 7 columns

# 选取第一个时间
# 用两种方法加六个月
first_date = crime_sort.index[0]
first_date
#Timestamp('2012-01-02 00:06:00')

first_date + pd.offsets.MonthBegin(6)
#Timestamp('2012-07-01 00:06:00')

first_date + pd.offsets.MonthEnd(6)
#Timestamp('2012-06-30 00:06:00')

更多

# 使用自定义的DateOffset对象
dt = pd.Timestamp('2012-1-16 13:40')
dt + pd.DateOffset(months=1)
#Timestamp('2012-02-16 13:40:00')

# 一个使用更多日期和时间的例子
do = pd.DateOffset(years=2, months=5, days=3, hours=8, seconds=10)
pd.Timestamp('2012-1-22 03:22') + do
#Timestamp('2014-06-25 11:22:10')

pd.options.display.max_rows=60