我有一个大的数据帧，我想在其中进行元素乘法，但只针对某些列。以下是DF 的示例
Name Age State Student_A Student_B Height_Student_A Height_Student_B
A 2 NZ 0.5 NA 0.5 0.2
B 1 AS 0.5 NA 0.5 0.2
C 4 MU NA 0.6 0.5 0.2
D 5 BY NA 0.7 0.5 0.2
目标是通过匹配的"高度"列将每个学生相乘。输出应该看起来像
Name Age State Student_A Student_B Height_Student_A Height_Student_B Score_Student_A Score_Student B
A 2 NZ 0.5 NA 0.5 0.2 0.25 NA
B 1 AS 0.5 NA 0.5 0.2 0.25 NA
C 4 MU NA 0.6 0.5 0.2 NA 0.12
D 5 BY NA 0.7 0.5 0.2 NA 0.14
我怀疑使用元素乘法，但我不确定如何指定从哪列开始。感谢您的帮助。
您可以捕获Student列和Height列，并直接将它们相乘。
student_cols <- sort(grep('^Student', names(df), value = TRUE))
height_cols <- sort(grep('^Height', names(df), value = TRUE))
df[paste0('Score_', student_cols)] <- df[student_cols] * df[height_cols]
df
# Name Age State Student_A Student_B Height_Student_A Height_Student_B Score_Student_A Score_Student_B
#1 A 2 NZ 0.5 NA 0.5 0.2 0.25 NA
#2 B 1 AS 0.5 NA 0.5 0.2 0.25 NA
#3 C 4 MU NA 0.6 0.5 0.2 NA 0.12
#4 D 5 BY NA 0.7 0.5 0.2 NA 0.14
游戏后期，但这可能有助于您(并解决@latlio的评论(将其推广到其他处理中(因为Ronak的交易是该数据的通用解决方案(。
稍微改变一下形状。不幸的是，我不是tidyr::pivot最强的，所以可能有更好的方法。
library(dplyr)
library(tidyr)
pivot_longer(dat, -c("Name", "Age", "State"), names_pattern = "(.*)_([^_]+)$", names_to = c("type", "AB")) %>%
pivot_wider(c(Name:State, AB), names_from = "type", values_from = "value") %>%
mutate(Score = Student * Height_Student)
# # A tibble: 8 x 7
# Name Age State AB Student Height_Student Score
# <chr> <int> <chr> <chr> <dbl> <dbl> <dbl>
# 1 A 2 NZ A 0.5 0.5 0.25
# 2 A 2 NZ B NA 0.2 NA
# 3 B 1 AS A 0.5 0.5 0.25
# 4 B 1 AS B NA 0.2 NA
# 5 C 4 MU A NA 0.5 NA
# 6 C 4 MU B 0.6 0.2 0.12
# 7 D 5 BY A NA 0.5 NA
# 8 D 5 BY B 0.7 0.2 0.140
由于R是矢量化的，您可以直接将列相乘以创建新列：
DF$Score_Student_A <- DF$Student_A * DF$Height_Student_A
DF$Score_Student_B <- DF$Student_B * DF$Height_Student_B