r - 当值首先超过阈值然后保持高于阈值时获取行

Question

我有以下数据框:

ID Date        Value
1  2010-08-01  6
1  2011-05-01  8
1  2011-12-01  7
1  2012-08-01  6
1  2013-01-01  6
1  2014-04-01  10
1  2014-08-01  8
1  2015-01-01  9
1  2016-01-01  9
1  2017-01-01  8
2  1996-01-01  5
2  1998-01-01  8
2  2010-01-01  5
2  2014-08-01  8
2  2016-01-01  8

我想为每个 ID 检测第一行 Value 等于或超过设定的阈值(在本例中: 8) 并且保持以下观察的阈值。因此，基于数据框，我想获得的结果如下:

ID Date       Value 
1 2014-04-01  10
2 2014-08-01  8

非常感谢您的帮助!谢谢!如果您需要任何其他信息或说明，请告诉我

Answer 1

不清楚 remains 情况 - 即假设 ID 的最后一个元素的“值”小于 8，则不清楚是否返回该 ID 的行。以下解决方案将跳过那些“ID”(如果存在)

library(dplyr)
library(data.table)
df1 %>% 
  group_by(ID) %>%
   slice({ind = Value >= 8
         grp = rleid(ind)
        if(all(ind[grp == max(grp)])) 
      first(row_number()[grp == max(grp)]) else 0
  }) %>% 
ungroup

-输出

# A tibble: 2 × 3
     ID Date       Value
  <int> <chr>      <int>
1     1 2014-04-01    10
2     2 2014-08-01     8

数据

df1 <- structure(list(ID = c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
2L, 2L, 2L, 2L, 2L), Date = c("2010-08-01", "2011-05-01", "2011-12-01", 
"2012-08-01", "2013-01-01", "2014-04-01", "2014-08-01", "2015-01-01", 
"2016-01-01", "2017-01-01", "1996-01-01", "1998-01-01", "2010-01-01", 
"2014-08-01", "2016-01-01"), Value = c(6L, 8L, 7L, 6L, 6L, 10L, 
8L, 9L, 9L, 8L, 5L, 8L, 5L, 8L, 8L)), class = "data.frame", row.names = c(NA, 
-15L))