dplyr 按组汇总大于阈值的总和值

Question

我有一个包含约 64,000 行的数据框，其结构如下

 df <- data.frame(FY = c(2018, 2018, 2019, 2019, 2020, 2020), WorkArea = c("03218", "03219", "07589", "07589", "01125", "01126"), ProjectCode = c(12, 12, 13, 11, 12, 12), TotalHours = c(180, 250, 125, 300, 450, 750))

我正在尝试为每年/工作区域/项目代码的每个分组创建一个大的汇总统计表：

• 操作次数
• 总时间
• 平均时间
• 平均时间
• 小时的第一个四分位数
• 小时的第三个四分位数
• 小时数>第 80 分位数的操作数量
• 运营总时间 > 第 80 分位数

我想我已经完成了最后一个...

df_table <- df %>% group_by(FY, WorkArea, ProjectCode) %>%
  summarise(opCount = n(),
            Hours = sum(TotalHours),
            AvgHours = Hours/opCount,
            medHours = median(TotalHours),
            Bottom25 = quantile(TotalHours, 0.25),
            Top25 = quantile(TotalHours, 0.75),
            OutlierCount = sum(TotalHours > quantile(TotalHours, 0.8))
            )

但我不知道如何有条件地对大于

quantile(Hours,0.8)

的小时数进行求和。

Answer 1

您可以在自定义摘要函数中使用

quantile

。

> fn <- \(x) {
+   q80 <- quantile(x, .8)
+   c(n=length(x), total=sum(x), summary(x), n80=length(x[x > q80]), total80=sum(x[x > q80]))
+ }
> res <- aggregate(TotalHours ~ WorkArea + ProjectCode, df1, fn)
> head(res)
  WorkArea ProjectCode TotalHours.n TotalHours.total TotalHours.Min.
1    03218          11       3.0000        1746.0000        398.0000
2    03219          11       3.0000        1330.0000        245.0000
3    07589          11       3.0000        1530.0000        397.0000
4    01125          11       3.0000        1317.0000        123.0000
5    01126          11       3.0000         944.0000        257.0000
6    03218          12       3.0000        1028.0000        123.0000
  TotalHours.1st Qu. TotalHours.Median TotalHours.Mean TotalHours.3rd Qu.
1           405.5000          413.0000        582.0000           674.0000
2           375.0000          505.0000        443.3333           542.5000
3           405.0000          413.0000        510.0000           566.5000
4           285.0000          447.0000        439.0000           597.0000
5           276.5000          296.0000        314.6667           343.5000
6           184.0000          245.0000        342.6667           452.5000
  TotalHours.Max. TotalHours.n80 TotalHours.total80
1        935.0000         1.0000           935.0000
2        580.0000         1.0000           580.0000
3        720.0000         1.0000           720.0000
4        747.0000         1.0000           747.0000
5        391.0000         1.0000           391.0000
6        660.0000         1.0000           660.0000

---

数据：

> dput(df1)
structure(list(WorkArea = structure(c(1L, 2L, 3L, 4L, 5L, 1L, 
2L, 3L, 4L, 5L, 1L, 2L, 3L, 4L, 5L, 1L, 2L, 3L, 4L, 5L, 1L, 2L, 
3L, 4L, 5L, 1L, 2L, 3L, 4L, 5L, 1L, 2L, 3L, 4L, 5L, 1L, 2L, 3L, 
4L, 5L, 1L, 2L, 3L, 4L, 5L), levels = c("03218", "03219", "07589", 
"01125", "01126"), class = "factor"), FY = c(2018, 2018, 2018, 
2018, 2018, 2019, 2019, 2019, 2019, 2019, 2020, 2020, 2020, 2020, 
2020, 2018, 2018, 2018, 2018, 2018, 2019, 2019, 2019, 2019, 2019, 
2020, 2020, 2020, 2020, 2020, 2018, 2018, 2018, 2018, 2018, 2019, 
2019, 2019, 2019, 2019, 2020, 2020, 2020, 2020, 2020), ProjectCode = c(12, 
12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 13, 13, 
13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 11, 11, 11, 
11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11), TotalHours = c(660L, 
420L, 252L, 173L, 327L, 245L, 733L, 148L, 227L, 402L, 123L, 938L, 
455L, 700L, 264L, 721L, 631L, 509L, 981L, 978L, 998L, 396L, 700L, 
382L, 720L, 616L, 311L, 959L, 971L, 358L, 413L, 580L, 397L, 123L, 
257L, 398L, 505L, 413L, 747L, 391L, 935L, 245L, 720L, 447L, 296L
)), out.attrs = list(dim = c(WorkArea = 5L, FY = 3L, ProjectCode = 3L
), dimnames = list(WorkArea = c("WorkArea=03218", "WorkArea=03219", 
"WorkArea=07589", "WorkArea=01125", "WorkArea=01126"), FY = c("FY=2018", 
"FY=2019", "FY=2020"), ProjectCode = c("ProjectCode=12", "ProjectCode=13", 
"ProjectCode=11"))), row.names = c(NA, -45L), class = "data.frame")