基于行内容的高度选择性过滤

Question

我有一个数据集（相当不整洁 - 但不是我的工作......帮助同事），其中有几行值，其中一些行与一列重复，但其他列由于添加到某些元素的“*”而不同。重复以下：-

a <- c("2020", "Rose", "r","r","s","s","i","i","r")
b <- c("2020", "Rose","r*","r*","s*","s*","s*","s*","s*")
c <- c("2020", "Lily","r","r","s","s","i","i","r")
d <- c("2020", "Tulip","r*","r*","r*","r*","s*","r*","r*")
e <- c("2020", "Tulip","s","s","r","s","s","r","r")

data <- rbind(a,b,c,d,e)

所以我的数据框看起来像这样...

  [,1]   [,2]    [,3] [,4] [,5] [,6] [,7] [,8] [,9]
a "2020" "Rose"  "r"  "r"  "s"  "s"  "i"  "i"  "r" 
b "2020" "Rose"  "r*" "r*" "s*" "s*" "s*" "s*" "s*"
c "2020" "Lily"  "r"  "r"  "s"  "s"  "i"  "i"  "r" 
d "2020" "Tulip" "r*" "r*" "r*" "r*" "s*" "r*" "r*"
e "2020" "Tulip" "s"  "s"  "r"  "s"  "s"  "r"  "r"

我需要删除第 2 列中重复的行（“Rose”、“Lily”等）并有选择地保留带有 * 的行，因此它看起来像这样...

  [,1]   [,2]    [,3] [,4] [,5] [,6] [,7] [,8] [,9]
b "2020" "Rose"  "r*" "r*" "s*" "s*" "s*" "s*" "s*"
c "2020" "Lily"  "r"  "r"  "s"  "s"  "i"  "i"  "r" 
d "2020" "Tulip" "r*" "r*" "r*" "r*" "s*" "r*" "r*"

我觉得与 lapply 捆绑在一起的函数可能是正确的方法，但不知道如何进行！！ - 任何想法

Answer 1

你可以试试这个。对于第二个条件 (*s)，它只检查第 3 列，因为它们似乎全部或 none.

tbl <- table( data[,2] )
rmv <- names( tbl[ tbl > 1 ] )

data[ !( data[,2] %in% rmv & !grepl("\*",data[,3])), ]
  [,1]   [,2]    [,3] [,4] [,5] [,6] [,7] [,8] [,9]
b "2020" "Rose"  "r*" "r*" "s*" "s*" "s*" "s*" "s*"
c "2020" "Lily"  "r"  "r"  "s"  "s"  "i"  "i"  "r"
d "2020" "Tulip" "r*" "r*" "r*" "r*" "s*" "r*" "r*"

万一它必须 select 基于任何 *（至少一个）使用这个

data[ !( data[,2] %in% rmv & apply( data[,3:9], 1, function(x) 
  any(!grepl("\*",x)) )), ]
  [,1]   [,2]    [,3] [,4] [,5] [,6] [,7] [,8] [,9]
b "2020" "Rose"  "r*" "r*" "s*" "s*" "s*" "s*" "s*"
c "2020" "Lily"  "r"  "r"  "s"  "s"  "i"  "i"  "r"
d "2020" "Tulip" "r*" "r*" "r*" "r*" "s*" "r*" "r*"

Answer 2

首先，您谈到了数据框，但到目前为止您使用的是矩阵。那我们先做一个数据框吧

df <- as.data.frame(data)

其次，我们可以使用 by()，它的工作原理与 lapply(split(x, g), FUN) 基本相同。作为拆分因子，我们使用前两列 1:2 并在每个切片上应用 grepl()。终于rbind().

df <- by(df, df[1:2], \(x) {
  if (nrow(x) > 1) {
    x[grepl('\*', x$V3), ]
  } else x}) |> (\(.) do.call(rbind, .))()

df
#     V1    V2 V3 V4 V5 V6 V7 V8 V9
# c 2020  Lily  r  r  s  s  i  i  r
# b 2020  Rose r* r* s* s* s* s* s*
# d 2020 Tulip r* r* r* r* s* r* r*

要清理行名称，请添加：

|> `rownames<-`(NULL)

注：R 版本 4.1.2 (2021-11-01)。

数据：

data <- structure(c("2020", "2020", "2020", "2020", "2020", "Rose", "Rose", 
"Lily", "Tulip", "Tulip", "r", "r*", "r", "r*", "s", "r", "r*", 
"r", "r*", "s", "s", "s*", "s", "r*", "r", "s", "s*", "s", "r*", 
"s", "i", "s*", "i", "s*", "s", "i", "s*", "i", "r*", "r", "r", 
"s*", "r", "r*", "r"), .Dim = c(5L, 9L), .Dimnames = list(c("a", 
"b", "c", "d", "e"), NULL))

基于行内容的高度选择性过滤

Highly selective filtering based on row contents

r

filtering

apply