將 R 分組 dataframe 中的所有行與當前行進行比較

Question

嗨，我正在嘗試確定組中是否有任何行的版本比該組的任何其他行小 1，並且 label 它在另一列中。 我已經查看了滯后和領先，但問題是值不同的 1 行可能彼此相鄰，也可能不相鄰。

這是一個可重現的示例

數據：

library(dplyr)

df <- tibble('Plate' = c("A1","A1","A1","A1","A1","A2","A2","A2","A2","A2","A3","A3","A3","A3","A3","A3"),
             'Sample' = c("a", "a","a","b","b","a","a","b","b","b","a","b","b","c","c","c"),
             'Location' = c("x","x","x","y","y","y","y","x","x","x","x","y","y","x","x","x"),
             'Version' = c(1,1.2,2,22,26,9,9.3,11,11.3,12,19,32.2,33.2,14,15,15))

我嘗試過的最后一次迭代改編自如何將當前行與 r（和其他）中的所有先前行進行比較

df_test <- df  %>%
  group_by(Plate,Sample,Location) %>% 
  arrange(desc(Version)) %>% 
  mutate(diff = sapply(seq_along(Version), function(i){
    if_else(any(.[1:(i-1),'Version'] - .[[i,'Version']] == 1.0), -1.0, 0)})
    )

預期 output：

   Plate Sample Location Version  diff
   <chr> <chr>  <chr>      <dbl> <dbl>
 1 A3    b      y           33.2     0
 2 A3    b      y           32.2    -1
 3 A1    b      y           26       0
 4 A1    b      y           22       0
 5 A3    a      x           19       0
 6 A3    c      x           15       0
 7 A3    c      x           15       0
 8 A3    c      x           14      -1
 9 A2    b      x           12       0
10 A2    b      x           11.3     0
11 A2    b      x           11      -1
12 A2    a      y            9.3     0
13 A2    a      y            9       0
14 A1    a      x            2       0
15 A1    a      x            1.2     0
16 A1    a      x            1      -1

實際 Output：

   Plate Sample Location Version  diff
   <chr> <chr>  <chr>      <dbl> <dbl>
 1 A3    b      y           33.2     0
 2 A3    b      y           32.2    -1
 3 A1    b      y           26       0
 4 A1    b      y           22      -1
 5 A3    a      x           19       0
 6 A3    c      x           15       0
 7 A3    c      x           15      -1
 8 A3    c      x           14       0
 9 A2    b      x           12       0
10 A2    b      x           11.3    -1
11 A2    b      x           11       0
12 A2    a      y            9.3     0
13 A2    a      y            9      -1
14 A1    a      x            2       0
15 A1    a      x            1.2    -1
16 A1    a      x            1       0

似乎正在查看行索引以進行比較（或忽略組？），我如何讓它查看價值？ 感覺就像我很接近。 我更喜歡 dplyr 答案，但如果需要，data.table 可以接受。 抱歉，如果我錯過了已經回答的相關帖子

Answer 1

我會試試這個：

df  %>%
  group_by(Plate,Sample,Location) %>%
  mutate(diff = if_else((Version + 1) %in% Version, -1, 0))
# # A tibble: 16 x 5
# # Groups:   Plate, Sample, Location [7]
#    Plate Sample Location Version  diff
#    <chr> <chr>  <chr>      <dbl> <dbl>
#  1 A1    a      x            1      -1
#  2 A1    a      x            1.2     0
#  3 A1    a      x            2       0
#  4 A1    b      y           22       0
#  5 A1    b      y           26       0
#  6 A2    a      y            9       0
#  7 A2    a      y            9.3     0
#  8 A2    b      x           11      -1
#  9 A2    b      x           11.3     0
# 10 A2    b      x           12       0
# 11 A3    a      x           19       0
# 12 A3    b      y           32.2    -1
# 13 A3    b      y           33.2     0
# 14 A3    c      x           14      -1
# 15 A3    c      x           15       0
# 16 A3    c      x           15       0

由於您的版本並非都是整數，因此存在一些數值精度問題的風險，但它似乎適用於您的數字相對較低的示例。

數值穩定的版本可能如下所示：

df  %>%
  group_by(Plate,Sample,Location) %>%
  mutate(diff = if_else(apply(abs(outer(Version, Version, "-") + 1) < 1e-10, 1, any), -1, 0))

（與上述結果相同）

要了解它的工作原理/方式，請從x = c(1, 1.2, 2)之類的文本向量開始，然后在其上運行代碼片段 - outer(x, x, "-") ，然后添加+ 1 ，等等