按R中的第一个和最后一个条件过滤行

Question

我想删除特定因素之前的所有行（在这种情况下是第一个 Running in Run 列）和最后一个特定情况之后的所有行（在这种情况下也是 Running in Run 列）。

数据

Date	Time	Run
17/04/12	00:10:00
17/04/12	00:19:00
17/04/12	00:25:00	Running
17/04/12	00:29:00	Running
17/04/12	00:25:00	Stop
17/04/12	00:29:00	Running
17/04/12	00:30:00	Running
18/04/12	00:10:00	Stop
18/04/12	00:14:00	Running
18/04/12	00:20:00	Stop
18/04/12	00:24:00	Stop
18/04/12	00:26:00	Stop

结果：

Date	Time	Run
17/04/12	00:25:00	Running
17/04/12	00:29:00	Running
17/04/12	00:25:00	Stop
17/04/12	00:29:00	Running
17/04/12	00:30:00	Running
18/04/12	00:10:00	Stop
18/04/12	00:14:00	Running

Answer 1

这是使用dplyr的解决方案

library(dplyr)

df %>% 
  filter(Run %in% c("Running", "Stop")) %>% 
  slice(1:max(which(Run == "Running")))

这给了我们：

# A tibble: 7 x 3
  Date     Time     Run    
  <chr>    <chr>    <chr>  
1 17/04/12 00:25:00 Running
2 17/04/12 00:29:00 Running
3 17/04/12 00:25:00 Stop   
4 17/04/12 00:29:00 Running
5 17/04/12 00:30:00 Running
6 18/04/12 00:10:00 Stop   
7 18/04/12 00:14:00 Running

Answer 2

在您的数据分析过程中，您会经常遇到这个问题。 如果您想要或将不得不添加更多逻辑检查，您可以引入您过滤的变量。在您的情况下，我们可以简单地计算 Running 和 Stop 的出现次数。然后，我们使用额外的列来过滤您的数据框。

library(dplyr)

data %>% 
 mutate(  RunCounts = cumsum(Run == "Running")
       , StopCounts = cumsum(Run == "Stop")
       ) %>% 
 filter(RunCounts >= 1 & StopCounts < max(StopCounts))

这会产生：

# A tibble: 7 x 5
  Date     Time     Run     RunCounts StopCounts
  <chr>    <chr>    <chr>       <int>      <int>
1 17/04/12 00:25:00 Running         1          2
2 17/04/12 00:29:00 Running         2          2
3 17/04/12 00:25:00 Stop            2          3
4 17/04/12 00:29:00 Running         3          3
5 17/04/12 00:30:00 Running         4          3
6 18/04/12 00:10:00 Stop            4          4
7 18/04/12 00:14:00 Running         5          4

您可以通过选择您感兴趣的列来删除引入的列：

data %>% mutate(RunCounts = cumsum(Run == "Running"), StopCounts = cumsum(Run == "Stop")) %>% filter(RunCounts >= 1 & StopCounts < max(StopCounts)) %>% select(Date, Time, Run)
# A tibble: 7 x 3
  Date     Time     Run    
  <chr>    <chr>    <chr>  
1 17/04/12 00:25:00 Running
2 17/04/12 00:29:00 Running
3 17/04/12 00:25:00 Stop   
4 17/04/12 00:29:00 Running
5 17/04/12 00:30:00 Running
6 18/04/12 00:10:00 Stop   
7 18/04/12 00:14:00 Running