R中的条件过滤和数据转换；根据组和值删除行并添加其他列

【问题标题】：Conditional filtering and data transformation in R; removing rows and adding additional columns based on group and valueR中的条件过滤和数据转换；根据组和值删除行并添加其他列
【发布时间】：2023-03-05 12:07:01
【问题描述】：

我正在尝试通过根据条件和分组过滤掉我不再需要的行来转换现有数据集。随后，我想添加一个附加列，它本质上是一个状态列和一个活动实例列。这是数据框：-

rawdata<-structure(list(DateTime = c("20/02/2021 13:00", "20/02/2021 14:00", 
                                     "20/02/2021 15:00", "20/02/2021 16:00", "20/02/2021 17:00", "20/02/2021 18:00", 
                                     "20/02/2021 19:00", "20/02/2021 20:00", "20/02/2021 21:00", "20/02/2021 22:00", 
                                     "20/02/2021 23:00", "21/02/2021 00:00", "01/03/2021 00:00", "01/03/2021 01:00", 
                                     "01/03/2021 02:00", "01/03/2021 03:00", "01/03/2021 04:00", "01/03/2021 05:00", 
                                     "01/03/2021 06:00", "01/03/2021 07:00", "01/03/2021 08:00", "01/03/2021 09:00", 
                                     "01/03/2021 10:00", "01/03/2021 11:00", "01/03/2021 12:00", "01/03/2021 13:00", 
                                     "20/02/2021 13:00", "20/02/2021 14:00", "20/02/2021 15:00", "20/02/2021 16:00", 
                                     "20/02/2021 17:00", "20/02/2021 18:00", "20/02/2021 19:00", "20/02/2021 20:00", 
                                     "20/02/2021 21:00", "20/02/2021 22:00"), Cluster = c("Cluster 3", 
                                                                                          "Cluster 3", "Cluster 3", "Cluster 3", "NotActive", "NotActive", 
                                                                                          "NotActive", "Cluster 2", "Cluster 1", "Cluster 3", "NotActive", 
                                                                                          "NotActive", "NotActive", "Cluster 5", "Cluster 5", "Cluster 4", 
                                                                                          "NotActive", "NotActive", "NotActive", "NotActive", "Cluster 2", 
                                                                                          "Cluster 2", "Cluster 3", "NotActive", "NotActive", "NotActive", 
                                                                                          "NotActive", "NotActive", "NotActive", "NotActive", "NotActive", 
                                                                                          "NotActive", "Cluster 1", "Cluster 2", "NotActive", "NotActive"
                                     ), UserID = c("AAA", "AAA", "AAA", "AAA", "AAA", "AAA", "AAA", 
                                                   "AAA", "AAA", "AAA", "AAA", "AAA", "AAA", "BBB", "BBB", "BBB", 
                                                   "BBB", "BBB", "BBB", "BBB", "BBB", "BBB", "BBB", "BBB", "BBB", 
                                                   "BBB", "DDD", "DDD", "DDD", "DDD", "DDD", "DDD", "DDD", "DDD", 
                                                   "DDD", "DDD")), class = "data.frame", row.names = c(NA, -36L))
print(rawdata)
           DateTime   Cluster UserID
           DateTime   Cluster UserID
1  20/02/2021 13:00 Cluster 3    AAA
2  20/02/2021 14:00 Cluster 3    AAA
3  20/02/2021 15:00 Cluster 3    AAA
4  20/02/2021 16:00 Cluster 3    AAA
5  20/02/2021 17:00 NotActive    AAA
6  20/02/2021 18:00 NotActive    AAA
7  20/02/2021 19:00 NotActive    AAA
8  20/02/2021 20:00 Cluster 2    AAA
9  20/02/2021 21:00 Cluster 1    AAA
10 20/02/2021 22:00 Cluster 3    AAA
11 20/02/2021 23:00 NotActive    AAA
12 21/02/2021 00:00 NotActive    AAA
13 01/03/2021 00:00 NotActive    AAA
14 01/03/2021 01:00 Cluster 5    BBB
15 01/03/2021 02:00 Cluster 5    BBB
16 01/03/2021 03:00 Cluster 4    BBB
17 01/03/2021 04:00 NotActive    BBB
18 01/03/2021 05:00 NotActive    BBB
19 01/03/2021 06:00 NotActive    BBB
20 01/03/2021 07:00 NotActive    BBB
21 01/03/2021 08:00 Cluster 2    BBB
22 01/03/2021 09:00 Cluster 2    BBB
23 01/03/2021 10:00 Cluster 3    BBB
24 01/03/2021 11:00 NotActive    BBB
25 01/03/2021 12:00 NotActive    BBB
26 01/03/2021 13:00 NotActive    BBB
27 20/02/2021 13:00 NotActive    DDD
28 20/02/2021 14:00 NotActive    DDD
29 20/02/2021 15:00 NotActive    DDD
30 20/02/2021 16:00 NotActive    DDD
31 20/02/2021 17:00 NotActive    DDD
32 20/02/2021 18:00 NotActive    DDD
33 20/02/2021 19:00 Cluster 1    DDD
34 20/02/2021 20:00 Cluster 2    DDD
35 20/02/2021 21:00 NotActive    DDD
36 20/02/2021 22:00 NotActive    DDD

为了进一步解释，这里是所需的输出：-

desiredoutput<-structure(list(DateTime = structure(c(1613826000, 1613829600, 
                                                     1613833200, 1613836800, 1613840400, 1613851200, 1613854800, 1613858400, 
                                                     1613862000, 1614560400, 1614564000, 1614567600, 1614571200, 1614585600, 
                                                     1614589200, 1614592800, 1614596400, 1613847600, 1613851200, 1613854800
), class = c("POSIXct", "POSIXt"), tzone = "UTC"), Cluster = c("Cluster 3", 
                                                               "Cluster 3", "Cluster 3", "Cluster 3", "NotActive", "Cluster 2", 
                                                               "Cluster 1", "Cluster 3", "NotActive", "Cluster 5", "Cluster 5", 
                                                               "Cluster 4", "NotActive", "Cluster 2", "Cluster 2", "Cluster 3", 
                                                               "NotActive", "Cluster 1", "Cluster 2", "NotActive"), UserID = c("AAA", 
                                                                                                                               "AAA", "AAA", "AAA", "AAA", "AAA", "AAA", "AAA", "AAA", "BBB", 
                                                                                                                               "BBB", "BBB", "BBB", "BBB", "BBB", "BBB", "BBB", "DDD", "DDD", 
                                                                                                                               "DDD"), Status = c("Start", "Ongoing", "Ongoing", "Ongoing", 
                                                                                                                                                  "Complete", "Start", "Ongoing", "Ongoing", "Complete", "Start", 
                                                                                                                                                  "Ongoing", "Ongoing", "Complete", "Start", "Ongoing", "Ongoing", 
                                                                                                                                                  "Complete", "Start", "Ongoing", "Complete"), Instance = c(1, 
                                                                                                                                                                                                            2, 3, 4, 5, 1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3)), row.names = c(NA, 
                                                                                                                                                                                                                                                                                     -20L), class = c("tbl_df", "tbl", "data.frame"))

print(desiredoutput)
      DateTime            Cluster   UserID Status   Instance
   <dttm>              <chr>     <chr>  <chr>       <dbl>
 1 2021-02-20 13:00:00 Cluster 3 AAA    Start           1
 2 2021-02-20 14:00:00 Cluster 3 AAA    Ongoing         2
 3 2021-02-20 15:00:00 Cluster 3 AAA    Ongoing         3
 4 2021-02-20 16:00:00 Cluster 3 AAA    Ongoing         4
 5 2021-02-20 17:00:00 NotActive AAA    Complete        5
 6 2021-02-20 20:00:00 Cluster 2 AAA    Start           1
 7 2021-02-20 21:00:00 Cluster 1 AAA    Ongoing         2
 8 2021-02-20 22:00:00 Cluster 3 AAA    Ongoing         3
 9 2021-02-20 23:00:00 NotActive AAA    Complete        4
10 2021-03-01 01:00:00 Cluster 5 BBB    Start           1
11 2021-03-01 02:00:00 Cluster 5 BBB    Ongoing         2
12 2021-03-01 03:00:00 Cluster 4 BBB    Ongoing         3
13 2021-03-01 04:00:00 NotActive BBB    Complete        4
14 2021-03-01 08:00:00 Cluster 2 BBB    Start           1
15 2021-03-01 09:00:00 Cluster 2 BBB    Ongoing         2
16 2021-03-01 10:00:00 Cluster 3 BBB    Ongoing         3
17 2021-03-01 11:00:00 NotActive BBB    Complete        4
18 2021-02-20 19:00:00 Cluster 1 DDD    Start           1
19 2021-02-20 20:00:00 Cluster 2 DDD    Ongoing         2
20 2021-02-20 21:00:00 NotActive DDD    Complete        3

我所追求的是在Cluster 列（由UserID）中，如果值为NotActive，则保留第一个NotActive 值并丢弃剩余的NotActive 行，直到它是别的东西.同样，我想创建一个“状态”列，其中该列中的第一个值对应于Start，直到第一个NotActive 值，它将对应于Complete； Ongoing 之间的所有其他内容。最后，Instance 列只是在Status 列中从 1==Start 到 N==Complete 编号。

任何帮助将不胜感激:)

【问题讨论】：

    标签： r if-statement dplyr conditional-statements conditional-formatting


    【解决方案1】：

    这是dplyr 方式-

    library(dplyr)

rawdata %>%
  #for each user id
  group_by(UserID) %>%
  #drop the repeated the 'NotActive' rows
  filter(Cluster != 'NotActive' | lag(Cluster != 'NotActive', default = TRUE)) %>%
  #Drop the rows that begins with 'NotActive'. 
  filter(!(row_number() == 1 & Cluster == 'NotActive')) %>%
  #Create a group column for each cycle
  group_by(grp = cumsum(lag(Cluster == 'NotActive', default = TRUE)), .add = TRUE) %>%
  #Create a status column which starts with 'Start' and ends with 'Complete'
  #fill it with 'Ongoing' in between. 
  #Create an Instance column using row_number()
  mutate(Status = c('Start', rep('Ongoing', n() - 2), 'Complete'), 
         Instance = row_number()) %>%
  ungroup %>%
  select(-grp)

    这将返回以下数据帧。

    #   DateTime         Cluster   UserID Status   Instance
#   <chr>            <chr>     <chr>  <chr>       <int>
# 1 20/02/2021 13:00 Cluster 3 AAA    Start           1
# 2 20/02/2021 14:00 Cluster 3 AAA    Ongoing         2
# 3 20/02/2021 15:00 Cluster 3 AAA    Ongoing         3
# 4 20/02/2021 16:00 Cluster 3 AAA    Ongoing         4
# 5 20/02/2021 17:00 NotActive AAA    Complete        5
# 6 20/02/2021 20:00 Cluster 2 AAA    Start           1
# 7 20/02/2021 21:00 Cluster 1 AAA    Ongoing         2
# 8 20/02/2021 22:00 Cluster 3 AAA    Ongoing         3
# 9 20/02/2021 23:00 NotActive AAA    Complete        4
#10 01/03/2021 01:00 Cluster 5 BBB    Start           1
#11 01/03/2021 02:00 Cluster 5 BBB    Ongoing         2
#12 01/03/2021 03:00 Cluster 4 BBB    Ongoing         3
#13 01/03/2021 04:00 NotActive BBB    Complete        4
#14 01/03/2021 08:00 Cluster 2 BBB    Start           1
#15 01/03/2021 09:00 Cluster 2 BBB    Ongoing         2
#16 01/03/2021 10:00 Cluster 3 BBB    Ongoing         3
#17 01/03/2021 11:00 NotActive BBB    Complete        4
#18 20/02/2021 19:00 Cluster 1 DDD    Start           1
#19 20/02/2021 20:00 Cluster 2 DDD    Ongoing         2
#20 20/02/2021 21:00 NotActive DDD    Complete        3

    【讨论】：

    • 我可以看到您的解决方案是如何工作的，但是，我发现我的数据有一些无法预料的事情。我可以简要更新我的 OP 以帮助说明吗？
    • 我已经添加了一个额外的例子。我有一些用户 ID 并不总是以“Cluster _”开头，而是以“NotActive”开头。请参阅我更新后的所需输出以获得更多说明，即用户 ID“DDD”。如果您能修改您的解决方案以考虑到这一点，将不胜感激:) 非常感谢您
    • @RobinTurkington 尝试更新的答案。
    • 太棒了，非常感谢！
    猜你喜欢
    • 1970-01-01
    • 2021-12-08
    • 2021-08-09
    • 2018-12-22
    • 2022-07-20
    • 2016-04-07
    • 2021-07-27
    • 1970-01-01
    • 1970-01-01
    相关资源
    最近更新 更多
    热门标签
    Java Python linux javascript Mysql C# Docker 算法 前端 SpringBoot Redis Vue spring 设计模式 .net core .net kubernetes c++ 数据库 数据结构 大数据 js 机器学习 微服务 Android Go 程序员 面试 JVM ASP.net core 云原生 人工智能 后端 PHP git CSS golang k8s Nginx Django mybatis 深度学习 多线程 React 架构 devops 爬虫 云计算 Spring Boot LeetCode