符号%>%，这是管道操作，其意思是将%>%左边的对象传递给右边的函数，作为第一个选项的设置(或剩下唯一一个选项的设置)
x %>%f(y) 等同于 f(x, y)
y %>%f(x, ., z) 等同于 f(x, y, z )
直接上栗子?：
library(data.table)library(tidyr)
# 先虚构一数据框
set.seed(1)
date
hour
min
second
event
data
head(data)[0:2]
## date hour min second event
## 1: 2016-11-01 7 30 29 u
## 2: 2016-11-02 9 43 36 a
typeof(data)
## [1] "list"
------------------------------------------管道函数(%>%)很简单-------------------------
# 把date，hour，min和second列合并为新列datetime
# R中的日期时间格式为"Year-Month-Day-Hour:Min:Second"
#(一)(%>%) 合着写
dataNew %unite(datehour, date, hour, sep = ' ') %>%unite(datetime, datehour, min, second, sep = ':')
dataNew
## datetime event
## 1: 2016-11-01 7:30:29 u
## 2: 2016-11-02 9:43:36 a
## 3: 2016-11-03 13:58:60 l
## 4: 2016-11-04 20:22:11 q
## 5: 2016-11-05 5:44:47 p
## 6: 2016-11-06 18:52:37 k
## 7: 2016-11-07 19:12:43 r
## 8: 2016-11-08 12:35:6 i
## 9: 2016-11-09 11:7:38 e
## 10: 2016-11-10 1:14:21 b
## 11: 2016-11-11 3:20:42 w
## 12: 2016-11-12 14:1:32 t
## 13: 2016-11-13 23:19:52 h
## 14: 2016-11-14 21:41:26 s
## 15: 2016-11-15 8:16:25 o
# #(二)(%>%) 分步写
dataNew %unite(datehour, date, hour, sep = ' '); dataNew
## datehour min second event
## 1: 2016-11-01 7 30 29 u
## 2: 2016-11-02 9 43 36 a
## 3: 2016-11-03 13 58 60 l
## 4: 2016-11-04 20 22 11 q
## 5: 2016-11-05 5 44 47 p
## 6: 2016-11-06 18 52 37 k
## 7: 2016-11-07 19 12 43 r
## 8: 2016-11-08 12 35 6 i
## 9: 2016-11-09 11 7 38 e
## 10: 2016-11-10 1 14 21 b
## 11: 2016-11-11 3 20 42 w
## 12: 2016-11-12 14 1 32 t
## 13: 2016-11-13 23 19 52 h
## 14: 2016-11-14 21 41 26 s
## 15: 2016-11-15 8 16 25 o
dataNew %>% unite(datetime, datehour, min, second, sep = ':')
## datetime event
## 1: 2016-11-01 7:30:29 u
## 2: 2016-11-02 9:43:36 a
## 3: 2016-11-03 13:58:60 l
## 4: 2016-11-04 20:22:11 q
## 5: 2016-11-05 5:44:47 p
## 6: 2016-11-06 18:52:37 k
## 7: 2016-11-07 19:12:43 r
## 8: 2016-11-08 12:35:6 i
## 9: 2016-11-09 11:7:38 e
## 10: 2016-11-10 1:14:21 b
## 11: 2016-11-11 3:20:42 w
## 12: 2016-11-12 14:1:32 t
## 13: 2016-11-13 23:19:52 h
## 14: 2016-11-14 21:41:26 s
## 15: 2016-11-15 8:16:25 o
# #(三)不用％>％也能写
dt
dt
## datehour min second event
## 1: 2016-11-01 7 30 29 u
## 2: 2016-11-02 9 43 36 a
## 3: 2016-11-03 13 58 60 l
## 4: 2016-11-04 20 22 11 q
## 5: 2016-11-05 5 44 47 p
## 6: 2016-11-06 18 52 37 k
## 7: 2016-11-07 19 12 43 r
## 8: 2016-11-08 12 35 6 i
## 9: 2016-11-09 11 7 38 e
## 10: 2016-11-10 1 14 21 b
## 11: 2016-11-11 3 20 42 w
## 12: 2016-11-12 14 1 32 t
## 13: 2016-11-13 23 19 52 h
## 14: 2016-11-14 21 41 26 s
## 15: 2016-11-15 8 16 25 o
dt2
dt2
## datatime event
## 1: 2016-11-01 7:30:29 u
## 2: 2016-11-02 9:43:36 a
## 3: 2016-11-03 13:58:60 l
## 4: 2016-11-04 20:22:11 q
## 5: 2016-11-05 5:44:47 p
## 6: 2016-11-06 18:52:37 k
## 7: 2016-11-07 19:12:43 r
## 8: 2016-11-08 12:35:6 i
## 9: 2016-11-09 11:7:38 e
## 10: 2016-11-10 1:14:21 b
## 11: 2016-11-11 3:20:42 w
## 12: 2016-11-12 14:1:32 t
## 13: 2016-11-13 23:19:52 h
## 14: 2016-11-14 21:41:26 s
## 15: 2016-11-15 8:16:25 o