如何通过函数data.table指示用于计算不同统计量的`n`

Question

数据框df1总结了不同人的一小时时间间隔。

df1<- data.frame(Round_datetime=c("2016-08-23 11:00:00","2016-08-23 11:00:00","2016-08-23 12:00:00","2016-08-23 12:00:00"),
                 Person= c("Sophie","Anna","Sophie","Anna"))
df1$Round_datetime<-as.POSIXct(df1$Round_datetime, format="%Y-%m-%d %H",tz="UTC")

df1

       Round_datetime Person
1 2016-08-23 11:00:00 Sophie
2 2016-08-23 11:00:00   Anna
3 2016-08-23 12:00:00 Sophie
4 2016-08-23 12:00:00   Anna

数据框df2会随时间提供有关这些人员的一些信息。

df2<- data.frame(DateTime=c("2016-08-23 10:29:08.324","2016-08-23 10:39:36.326","2016-08-23 10:44:08.724","2016-08-23 10:59:46.324","2016-08-23 11:19:22.324","2016-08-23 11:29:53.324","2016-08-23 11:34:14.324","2016-08-23 11:47:49.324","2016-08-23 11:54:58.324","2016-08-23 11:59:13.324","2016-08-23 12:12:34.324","2016-08-23 12:23:43.324","2016-08-23 12:32:14.324","2016-08-23 12:29:28.324"),
                 Person=c("Sophie","Anna","Sophie","Anna","Sophie","Anna","Sophie","Anna","Sophie","Anna","Sophie","Anna","Sophie","Anna"),
                 Value=c(10,15,5,10,20,15,10,5,25,15,10,5,10,20))
df2$DateTime<-as.POSIXct(df2$DateTime, format="%Y-%m-%d %H:%M:%OS",tz="UTC")

df2

                  DateTime Person Value
1  2016-08-23 10:29:08.323 Sophie    10
2  2016-08-23 10:39:36.325   Anna    15
3  2016-08-23 10:44:08.723 Sophie     5
4  2016-08-23 10:59:46.323   Anna    10
5  2016-08-23 11:19:22.323 Sophie    20
6  2016-08-23 11:29:53.323   Anna    15
7  2016-08-23 11:34:14.323 Sophie    10
8  2016-08-23 11:47:49.323   Anna     5
9  2016-08-23 11:54:58.323 Sophie    25
10 2016-08-23 11:59:13.323   Anna    15
11 2016-08-23 12:12:34.323 Sophie    10
12 2016-08-23 12:23:43.323   Anna     5
13 2016-08-23 12:32:14.323 Sophie    10
14 2016-08-23 12:29:28.323   Anna    20

我用如下所示的代码添加的统计数据mean，standard deviation并standard error以df1根据该信息df2。

library(plotrix)
setDT(df1)[, Round_datetime := ymd_hms(Round_datetime)]
setDT(df2)[, dt_floor := round_date(ymd_hms(DateTime), unit = "hour")]
df2[df1, .(mean = mean(Value), 
          sd = sd(Value),
          se = std.error(Value)),
on = .(Person, dt_floor = Round_datetime), by = .EACHI]

   Person            dt_floor     mean        sd        se
1: Sophie 2016-08-23 11:00:00 12.50000 10.606602       7.49
2:   Anna 2016-08-23 11:00:00 13.33333  2.886751       1.66
3: Sophie 2016-08-23 12:00:00 15.00000  8.660254       4.99
4:   Anna 2016-08-23 12:00:00 11.25000  7.500000       3.75

但是，我将需要包含另一个名为的变量n，该变量指示每个一小时的时间间隔采集的样本数量。我期望的是：

   Person            dt_floor     mean        sd        se     n
1: Sophie 2016-08-23 11:00:00 12.50000 10.606602       7.49    2
2:   Anna 2016-08-23 11:00:00 13.33333  2.886751       1.66    3
3: Sophie 2016-08-23 12:00:00 15.00000  8.660254       4.99    3
4:   Anna 2016-08-23 12:00:00 11.25000  7.500000       3.75    4

有人知道怎么做吗？

Answer 1

只需添加.N到最后一部分：

df2[df1, .(mean = mean(Value), 
           sd = sd(Value),
           n = .N),
    on = .(Person, dt_floor = Round_datetime), by = .EACHI]

输出：

   Person            dt_floor     mean        sd n
1: Sophie 2016-08-23 11:00:00 12.50000 10.606602 2
2:   Anna 2016-08-23 11:00:00 13.33333  2.886751 3
3: Sophie 2016-08-23 12:00:00 15.00000  8.660254 3
4:   Anna 2016-08-23 12:00:00 11.25000  7.500000 4