r - 如何为另一列中定义的每种类型规范化一列中的值

Question

BEFORE
-------------------------------
ID            measure     value
-------------------------------
original      weight      120.0
variant1      weight      110.0
variant2      weight       78.0
variant3      weight      140.0
original      speed        56.0
variant1      speed        54.0
variant2      speed        56.0
variant3      speed        61.0
original      height        6.7
variant1      height        6.3
variant2      height        4.5
variant3      height        5.3
-------------------------------

假设我有一张那样的 table 。对于每种类型的“度量”，我想用在“原始”中找到的值对“值”进行标准化。这样结果看起来像。附加名为“norm_value”的新列

AFTER
------------------------------------
ID            measure     norm_value
------------------------------------
original      weight       1.0
variant1      weight       0.91
variant2      weight       0.65
variant3      weight       1.16
original      speed        1.0
variant1      speed        0.96
variant2      speed        1.0
variant3      speed        1.08
original      height       1.0
variant1      height       0.94
variant2      height       0.67
variant3      height       0.79
-------------------------------

非常感谢任何帮助。

Answer 1

使用dplyr，你可以尝试:

df %>%
 group_by(measure) %>%
 mutate(norm_value = value/value[ID == "original"])

   ID       measure value norm_value
   <chr>    <chr>   <dbl>      <dbl>
 1 original weight  120        1    
 2 variant1 weight  110        0.917
 3 variant2 weight   78        0.65 
 4 variant3 weight  140        1.17 
 5 original speed    56        1    
 6 variant1 speed    54        0.964
 7 variant2 speed    56        1    
 8 variant3 speed    61        1.09 
 9 original height    6.7      1    
10 variant1 height    6.3      0.940
11 variant2 height    4.5      0.672
12 variant3 height    5.3      0.791

与 base R 相同的想法:

with(df, value/ave((ID == "original") * value, measure, FUN = max))