Question

我有许多连续变量列，我需要计算其中有多少列在 x 范围内。我尝试了一种类似于此的相当冗长的方法：

library(tidyverse)

mtcars %>% 
  mutate("1" = if_else(mpg  >= 0 & mpg  <= 1, 1, 0),
         "5" = if_else(mpg  >= 0 & mpg  <= 5, 1, 0),
         "10" = if_else(mpg  >= 0 & mpg  <= 10, 1, 0),
         "20" = if_else(mpg  >= 0 & mpg  <= 20, 1, 0),
         "50" = if_else(mpg  >= 0 & mpg  <= 50, 1, 0),
         "100" = if_else(mpg  >= 0 & mpg  <= 100, 1, 0),
         "400" = if_else(mpg  >= 0 & mpg  <= 400, 1, 0),
         distance = round(mpg , 2)) %>% 
  select("1", "5", "10", "20", "50", "100", "400") %>% 
  map_df(mean)

有没有一种优雅的方法来迭代这个过程而不复制和粘贴变量？

Answer 1

如果我们需要对所有数字列执行此操作，则循环上范围值，然后使用 summarise 和 across 循环跨列，计算 mean使用 between

创建的逻辑向量

library(dplyr)
library(purrr)
imap_dfr(lst(1, 5, 10, 20, 50, 100, 400), ~ {
       ul <- .x
       mtcars %>% 
         summarise(across(where(is.numeric),
            ~ mean(between(., 0, ul))))}, .id = 'categ')

-输出

#  categ    mpg     cyl    disp      hp drat      wt    qsec vs am gear    carb
#1     1 0.0000 0.00000 0.00000 0.00000    0 0.00000 0.00000  1  1    0 0.21875
#2     5 0.0000 0.34375 0.00000 0.00000    1 0.90625 0.00000  1  1    1 0.93750
#3    10 0.0000 1.00000 0.00000 0.00000    1 1.00000 0.00000  1  1    1 1.00000
#4    20 0.5625 1.00000 0.00000 0.00000    1 1.00000 0.90625  1  1    1 1.00000
#5    50 1.0000 1.00000 0.00000 0.00000    1 1.00000 1.00000  1  1    1 1.00000
#6   100 1.0000 1.00000 0.15625 0.28125    1 1.00000 1.00000  1  1    1 1.00000
#7   400 1.0000 1.00000 0.90625 1.00000    1 1.00000 1.00000  1  1    1 1.00000

或者这可以在 base R 中使用 outer

完成

categ <- c(1, 5, 10, 20, 50, 100, 400)
out <- cbind(categ, outer(categ, mtcars,
     Vectorize(function(x, y) mean(y >= 0 & y <= x))))

-输出

out
#     categ    mpg     cyl    disp      hp drat      wt    qsec vs am gear    carb
#[1,]     1 0.0000 0.00000 0.00000 0.00000    0 0.00000 0.00000  1  1    0 0.21875
#[2,]     5 0.0000 0.34375 0.00000 0.00000    1 0.90625 0.00000  1  1    1 0.93750
#[3,]    10 0.0000 1.00000 0.00000 0.00000    1 1.00000 0.00000  1  1    1 1.00000
#[4,]    20 0.5625 1.00000 0.00000 0.00000    1 1.00000 0.90625  1  1    1 1.00000
#[5,]    50 1.0000 1.00000 0.00000 0.00000    1 1.00000 1.00000  1  1    1 1.00000
#[6,]   100 1.0000 1.00000 0.15625 0.28125    1 1.00000 1.00000  1  1    1 1.00000
#[7,]   400 1.0000 1.00000 0.90625 1.00000    1 1.00000 1.00000  1  1    1 1.00000

Answer 2

这应该给出相同的结果，并且我认为这是一种相当“整洁”的方式：

distance_thresholds <- tibble(
  threshold = c(1, 5, 10, 20, 50, 100, 400)
)

mtcars %>%
  left_join(distance_thresholds, by = character()) %>%
  group_by(threshold) %>%
  summarise(avg = mean(mpg <= threshold)) %>%
  pivot_wider(names_from = threshold, values_from = avg)

Answer 3

sapply 和 colMeans :

vals <- c(1, 5, 10, 20, 50, 100, 400)
colMeans(sapply(vals, function(x) mtcars$mpg >=0 & mtcars$mpg <= x))

计算许多变量的 x 范围的平均值

3 个答案: