Question

我有一个数据框，其中每一列都有值。有没有一种方法可以创建新列并查看“最大值”列的顺序。例如。请参考我的预期输出

df
COlA    COlB    COLC    COLD
 34      40      4.5      50
 35      70      2.0      30
 90      40      4.0      10

预期产量

df
COlA    COlB    COLC    COLD            NewCOl
 34      40      4.5      50      COLD  > COLB > COLA > COLC
 35      70      2.0      30      COLB > COLA > COLD > COLC
 90      40      4.0      10      COLA > COLB  > COLD > COLC

Answer 1

在基本R中，您可以这样操作（仅在没有联系的情况下才是正确的解决方案）：

dat$NewCol <- by(
  unlist(dat),
  row(dat),
  function(x) paste(names(dat)[order(x, decreasing = T)], collapse = ' > ')
  )

dat

#   COlA COlB COLC COLD                    NewCol
# 1   34   40  4.5   50 COLD > COlB > COlA > COLC
# 2   35   70  2.0   30 COlB > COlA > COLD > COLC
# 3   90   40  4.0   10 COlA > COlB > COLD > COLC

数据：

dat <- structure(
  list(
    COlA = c(34L, 35L, 90L),
    COlB = c(40L, 70L, 40L),
    COLC = c(4.5, 2, 4),
    COLD = c(50L, 30L, 10L)
  ),
  class = "data.frame",
  row.names = c(NA,-3L)
)

Answer 2

我们可以在创建行号列之后将其重塑为“长”格式，然后通过基于“值”列中的递减值paste将“名称”列ordered设置为新列，重新调整为“宽”格式，并select列

library(dplyr)
library(tidyr) #v 1.0.0    
library(stringr)
df %>%
   mutate(rn = row_number()) %>%
   pivot_longer(cols = -rn) %>%
   # or use gather for older versions
   # gather(name, value, -rn) %>%       
   group_by(rn) %>% 
   mutate(NewCol = str_c(name[order(-value)], collapse=' > ')) %>% 
   pivot_wider(names_from = name, values_from = value) %>%
   # or use spread for older versions
   # spread(name, value) %>%
   ungroup %>%
   select(names(df), NewCol)
# A tibble: 3 x 5
#   COlA  COlB  COLC  COLD NewCol                   
#  <dbl> <dbl> <dbl> <dbl> <chr>                    
#1    34    40   4.5    50 COLD > COlB > COlA > COLC
#2    35    70   2      30 COlB > COlA > COLD > COLC
#3    90    40   4      10 COlA > COlB > COLD > COLC

或者另一个选择是pmap

library(purrr)
df %>% 
   mutate(NewCol =pmap_chr(., ~ c(...) %>%
            {names(.)[order(-.)]} %>%
            str_c(collapse=" > ")))

如果有character列，请使用select_if

df %>%
     mutate(NewCol = pmap_chr(select_if(., is.numeric), ~ c(...) %>%
              {names(.)[order(-.)]} %>%
            str_c(collapse=" > ")))

数据

df <-  structure(list(COlA = c(34L, 35L, 90L), COlB = c(40L, 70L, 40L
), COLC = c(4.5, 2, 4), COLD = c(50L, 30L, 10L)),
class = "data.frame", row.names = c(NA, 
-3L))

查找具有最大值的列的顺序

2 个答案:

数据