基于数据帧R中排名列的数据帧子集列表

Question

我有一个数据帧列表。我只想对包含分数比第二个排名分数低 10 倍的行的数据帧进行子集化，删除所有其他数据帧。知道如何处理这个吗？谢谢！

>Output
$E1
  ID    model   score
  E1      AAA    2
  E1      BBB    100
  E1      CCC    130
  E1      ZZZ    120
  E1      YYY    128

$E2
  ID    model   score
  E2      XXX    130
  E2      ASD    144
  E2      DFE    142
  E2      FGS    145
  E2      GFH    124

首选结果：

>Output_subset
$E1
  ID    model   score
  E1      AAA    2
  E1      BBB    100
  E1      CCC    130
  E1      ZZZ    120
  E1      YYY    128

Answer 1

您可以编写一个函数来检查两个分数之间的条件：

check_data <- function(df) {
   x <- sort(df$score)
   x[1] < (x[2]/10)
}

您可以在 Filter 中使用此函数 in base R :

Filter(check_data, Output)

#$E1
#  ID model score
#1 E1   AAA     2
#2 E1   BBB   100
#3 E1   CCC   130
#4 E1   ZZZ   120
#5 E1   YYY   128

或 keep 在 purrr 中：

purrr::keep(Output, check_data)

数据

Output <- list(E1 = structure(list(ID = c("E1", "E1", "E1", "E1", "E1"), 
model = c("AAA", "BBB", "CCC", "ZZZ", "YYY"), score = c(2L, 
100L, 130L, 120L, 128L)), class = "data.frame", row.names = c(NA, 
-5L)), E2 = structure(list(ID = c("E2", "E2", "E2", "E2", "E2"
), model = c("XXX", "ASD", "DFE", "FGS", "GFH"), score = c(130L, 
144L, 142L, 145L, 124L)), class = "data.frame", row.names = c(NA, -5L)))

Answer 2

我们还可以使用 base R

中的 sapply

Output[sapply(Output, function(x) 
      with(head(x[order(x$score), ], 2), score[1] < (score[2]/10)))]

数据

Output <- list(E1 = structure(list(ID = c("E1", "E1", "E1", "E1", "E1"), 
model = c("AAA", "BBB", "CCC", "ZZZ", "YYY"), score = c(2L, 
100L, 130L, 120L, 128L)), class = "data.frame", row.names = c(NA, 
-5L)), E2 = structure(list(ID = c("E2", "E2", "E2", "E2", "E2"
), model = c("XXX", "ASD", "DFE", "FGS", "GFH"), score = c(130L, 
144L, 142L, 145L, 124L)), class = "data.frame", row.names = c(NA, -5L)))

基于数据帧R中排名列的数据帧子集列表

Subsetting list of dataframes based on ranked column in dataframes R

r

list

subset

ranking

dataframe

数据