R bind_rows 在使用 %>% 时表现不一致

Question

我有以下小标题

tst <- tibble(
  x = 'actual data',
  age_1 = 5.3,
  age_2 = 6.6,
  age_3 = 8.3,
  age_4 = 20.3,
  age_5 = 25.3,
  age_6 = 30.8,
  age_7 = 31.3,
  age_8 = 22.3,
  age_9 = 18.3,
  age_10 = 14.3
)

我可以用

创建新的第一行四舍五入值行

demo <- tst %>% 
  c(x='round',round(.[nrow(.),2:(ncol(.))])) %>% 
  bind_rows(tst,.)

demo
# A tibble: 2 x 11
  x           age_1 age_2 age_3 age_4 age_5 age_6 age_7 age_8 age_9 age_10
  <chr>       <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>  <dbl>
1 actual data  5.30  6.60  8.30  20.3  25.3  30.8  31.3  22.3  18.3   14.3
2 round        5.00  7.00  8.00  20.0  25.0  31.0  31.0  22.0  18.0   14.0

现在，当我尝试使用相同的代码创建另一行底值时，出现错误

demo %>%
  c(x='floor',round(demo[1,2:(ncol(demo))])) %>% 
  bind_rows(demo,.)
Error in bind_rows_(x, .id) : Argument 12 must be length 2, not 1

但是，如果我以更圆的方式做我认为相同的事情，它会起作用

i <- c(x='floor',round(demo[1,2:(ncol(demo))]))
bind_rows(demo,i)
# A tibble: 3 x 11
  x           age_1 age_2 age_3 age_4 age_5 age_6 age_7 age_8 age_9 age_10
  <chr>       <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>  <dbl>
1 actual data  5.30  6.60  8.30  20.3  25.3  30.8  31.3  22.3  18.3   14.3
2 round        5.00  7.00  8.00  20.0  25.0  31.0  31.0  22.0  18.0   14.0
3 floor        5.00  7.00  8.00  20.0  25.0  31.0  31.0  22.0  18.0   14.0

我不知道为什么当运行与我创建圆角行时的代码基本相同时会出现该错误。有任何想法吗？如果您有更好的建议（根据上面的行添加行），我很乐意听取他们的意见。

Answer 1

在 R 中处理宽数据集可能非常违反直觉。以长格式处理数据几乎总是更可取。直到您可能想要以宽 table 显示的程度。

我会像这样重塑您的数据：

library(tidyr)
library(dplyr)

tst <- tibble(
  x = 'actual data',
  age_1 = 5.3,
  age_2 = 6.6,
  age_3 = 8.3,
  age_4 = 20.3,
  age_5 = 25.3,
  age_6 = 30.8,
  age_7 = 31.3,
  age_8 = 22.3,
  age_9 = 18.3,
  age_10 = 14.3
)

df <- tst %>%
  select(-x) %>%
  gather(var, actual) %>%
  mutate(
    var = "age",
    round = round(actual),
    floor = floor(actual)
  )

df
# # A tibble: 10 x 4
#  var   actual round floor
#  <chr>  <dbl> <dbl> <dbl>
#  1 age   5.30  5.00  5.00
#  2 age   6.60  7.00  6.00
#  3 age   8.30  8.00  8.00
#  4 age  20.3  20.0  20.0 
#  5 age  25.3  25.0  25.0 
#  6 age  30.8  31.0  30.0 
#  7 age  31.3  31.0  31.0 
#  8 age  22.3  22.0  22.0 
#  9 age  18.3  18.0  18.0 
# 10 age  14.3  14.0  14.0

您现在可以看到添加 round 和 floor 是多么容易。要从控制台以宽格式快速显示，您可以将其转换为原始问题的表示形式。

df %>% select(-var) %>% t

#        [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10]
# actual  5.3  6.6  8.3 20.3 25.3 30.8 31.3 22.3 18.3  14.3
# round   5.0  7.0  8.0 20.0 25.0 31.0 31.0 22.0 18.0  14.0
# floor   5.0  6.0  8.0 20.0 25.0 30.0 31.0 22.0 18.0  14.0

一种 tidy 方法，而不是快速查看可能是：

df %>%
  mutate(cols = paste(var, sprintf("%02d", seq_len(nrow(.))), sep = "_")) %>%
  gather(var, value, -cols) %>%
  filter(var != "var") %>%
  spread(cols, value) %>%
  mutate_at(vars(2:length(.)), as.numeric)

# # A tibble: 3 x 11
#   var    age_01 age_02 age_03 age_04 age_05 age_06 age_07 age_08 age_09 age_10
#   <chr>   <dbl>  <dbl>  <dbl>  <dbl>  <dbl>  <dbl>  <dbl>  <dbl>  <dbl>  <dbl>
# 1 actual   5.30   6.60   8.30   20.3   25.3   30.8   31.3   22.3   18.3   14.3
# 2 floor    5.00   6.00   8.00   20.0   25.0   30.0   31.0   22.0   18.0   14.0
# 3 round    5.00   7.00   8.00   20.0   25.0   31.0   31.0   22.0   18.0   14.0

N.B. In this variant, I added the mutate_at because the numeric values were coerced to character by the gather of mixed types.

R bind_rows 在使用 %>% 时表现不一致

R bind_rows not behaving consistently when using %>%

r

tidyverse