R - 根据 group_id 和条件语句创建新索引

Question

我正在使用如下所示的数据框（让我们调用 MyData）。我想做的是按 PatientKey 分组并创建一个名为 NewID 的新 ID。每次对于 TimeBetweenTests > 14 的同一个 PatientKey，新的 Id 应该增加 1，并保持在该特定的新值上，直到出现新的 PatientKey 或对于相同的 PatientKey，出现新的 TimeBetweenTests > 14。

PatientKey             TimeBetweenTests     NewId        
1                      0                    NewId should be 1 (first patient)
1                      0                    NewId should be 1
1                      1                    NewId should be 1                                                                
1                      2                    NewId should be 1
2                      3                    NewId should be 2 (new patient)                                                                          
3                      4                    NewId should be 3 (new patient)      
3                      16                   NewId should be 4 (same patient but TimeBetweenTests > 14)                                                                                              
3                      80                   NewId should be 5 (same patient but TimeBetweenTests > 14)
4                      3                    NewId should be 6 (new patient)
4                      0                    NewId should be 6 (new patient)                                                                            
4                      90                   NewId should be 7 (same patient but TimeBetweenTests > 14)        
4                      110                  NewId should be 8 (same patient but TimeBetweenTests > 14) 
5                      3                    NewId should be 9 (new patient)
5                      3                    NewId should be 9
5                      3                    NewId should be 9

etc    
                                                                       

我已尝试为此使用 dplyr，但问题是当我尝试类似以下代码时后续值不会改变：

MyData % group_by(PatientKey) %>% mutate(NewId = ifelse(TimeBetweenTests > 14, lag(NewId), NewId))

任何人都有一个方便的 dplyr 或 data.table 解决方案，或者 for 循环方法。

Answer 1

试试这个

library(dplyr)
df %>% mutate(NewID = cumsum(lag(PatientKey, default = 0) != PatientKey | TimeBetweenTests > 14)

输出

   PatientKey TimeBetweenTests NewID
        <dbl>            <dbl> <int>
 1          1                0     1
 2          1                0     1
 3          1                1     1
 4          1                2     1
 5          2                3     2
 6          3                4     3
 7          3               16     4
 8          3               80     5
 9          4                3     6
10          4                0     6
11          4               90     7
12          4              110     8
13          5                3     9
14          5                3     9
15          5                3     9