Question

我有下表：

Class  x2  x3  x4
  A    14  45  53 
  A     8  18  17
  A    16  49  20
  B    78  21  48 
  B     8  18   5

我需要每个＆＃34; Class＆＃34; （A和B）在列＆＃34; X3＆＃34;中找到最大值，保留该行并删除其他行。

输出应采用以下格式：

Class  x2  x3  x4
  A    14  49  20
  B    78  21  48

如果我的问题不清楚，请问我问题。

谢谢！

Answer 1

基础R方法可能是：

mydf[as.logical(with(mydf, ave(x3, Class, FUN = function(x) x == max(x)))), ]
#   Class x2 x3 x4
# 3     A 16 49 20
# 4     B 78 21 48

但请注意，如果max绑定了多个值，则会为该组返回多行。

这是一个可能的＆＃34; data.table＆＃34;的方法：

library(data.table)
setkey(as.data.table(mydf), Class, x3)[, tail(.SD, 1), by = Class]
#    Class x2 x3 x4
# 1:     A 16 49 20
# 2:     B 78 21 48

Answer 2

使用dplyr的一种方法是：

library(dplyr)

foo %>%
    #For each Class
    group_by(Class) %>%
    # Sort rows in descending way using x3: you get the max x3 value on top
    # for each group
    arrange(desc(x3)) %>%
    # Select the first row for each Class
    slice(1)

#  Class x2 x3 x4
#1     A 16 49 20
#2     B 78 21 48

修改鉴于@Ananda的领带价值考虑和他在演出中的建议，你也可以这样做。但是，@ Richard Acriven的想法是如果有关系，那就该走了。

# Data foo2 <- structure(list(Class = structure(c(1L, 1L, 1L, 2L, 2L), .Label = c("A", "B"), class = "factor"), x2 = c(14L, 8L, 16L, 78L, 8L), x3 = c(49L, 18L, 49L, 21L, 18L), x4 = c(53L, 17L, 20L, 48L, 5L)), .Names = c("Class", "x2", "x3", "x4"), class = "data.frame", row.names = c(NA, -5L )) # Class x2 x3 x4 #1 A 14 49 53 #2 A 8 18 17 #3 A 16 49 20 #4 B 78 21 48 #5 B 8 18 5 foo2 %>% group_by(Class) %>% mutate(Rank = dense_rank(desc(x3))) %>% filter(Rank == 1) # Class x2 x3 x4 Rank #1 A 14 49 53 1 #2 A 16 49 20 1 #3 B 78 21 48 1

Answer 3

这是该批次的另一个dplyr答案

library(dplyr)
df %>% group_by(Class) %>% filter(x3 == max(x3))
# Source: local data frame [2 x 4]
# Groups: Class
# 
#   Class x2 x3 x4
# 1     A 16 49 20
# 2     B 78 21 48

也可能是

group_by(df, Class) %>% filter(x3 == max(x3))

Answer 4

尝试：

do.call(rbind, lapply(split(ddf, ddf$Class), function(x) tail(x[order(x$x3),],1)))
  Class x2 x3 x4
A     A 16 49 20
B     B 78 21 48

条件计算列中的最大值

4 个答案: