Multiply multiple columns (e.g. 1A*1B and 2A*2B) - r

Question

I'm looking for a fast way to multipy multiple columns with a specific ending.
I have this table
Revenue_1
Revenue_2
Sales_1
Sales_2
2,000
1,000
100
50
2,500
4,000
80
200
I want these additional columns by multiplying Revenue_N with Sales_N - For example, multiplying Revenue_1 * Sales_1 = 2,000 * 100 = 200,000
Revenue_1
Revenue_2
Sales_1
Sales_2
Sales_1
Sales_2
2,000
1,000
100
50
200,000
50,000
2,500
4,000
50
200
125,000
800,000
I was playing around with for-loops or using some form of apply but coulnd't figure it out. Any help?

Answer 1

Assuming dat as given reproducibly in the Note at the end we can just multiply the first two columns by the last two columns like this in base R. Note that in the question that the new columns have the same names as other columns which is not a good idea since then you can't reference them by name so here we have used new names. You could always change the names if you don't like the ones we used.
If there were additional columns that should not participate then instead of using -ix we could use grep("Sales", names(dat)) .
We computed ix but it would be possible to simply hard code it as (1:2) and then omit the first line.
ix <- grep("Revenue", names(dat)) # 1:2
cbind(dat, prod = dat[ix] * dat[-ix])
## Revenue_1 Revenue_2 Sales_1 Sales_2 prod.Revenue_1 prod.Revenue_2
## 1 2000 1000 100 50 2e+05 5e+04
## 2 2500 4000 80 200 2e+05 8e+05
Note
dat <- structure(list(Revenue_1 = c(2000, 2500), Revenue_2 = c(1000,
4000), Sales_1 = c(100, 80), Sales_2 = c(50, 200)), row.names = c(NA,
-2L), class = "data.frame")

Answer 2

We could also do this with across in mutate
library(dplyr)
library(stringr)
df1 <- df %>%
mutate(across(starts_with('Sales'), ~ .* get(str_replace(cur_column(), 'Sales',
'Revenue')), .names = 'Revenue_{.col}'))
-output
df1
# Revenue_1 Revenue_2 Sales_1 Sales_2 Revenue_Sales_1 Revenue_Sales_2
#1 2000 1000 100 50 200000 50000
#2 2500 4000 80 200 200000 800000
data
df <- structure(list(Revenue_1 = c(2000L, 2500L), Revenue_2 = c(1000L,
4000L), Sales_1 = c(100L, 80L), Sales_2 = c(50L, 200L)),
class = "data.frame", row.names = c(NA,
-2L))

Answer 3

Try the base R option below
> cbind(df, do.call("*", rev(split.default(df, gsub("\\d", "", names(df))))))
Revenue_1 Revenue_2 Sales_1 Sales_2 Sales_1 Sales_2
1 2000 1000 100 50 200000 50000
2 2500 4000 80 200 200000 800000
Data
> dput(df)
structure(list(Revenue_1 = c(2000L, 2500L), Revenue_2 = c(1000L,
4000L), Sales_1 = c(100L, 80L), Sales_2 = c(50L, 200L)), class = "data.frame", row.names = c(NA,
-2L))

Answer 4

I think this is what you asked for
df <- data.frame(Revenue_1=c(2000,2500),Revenue_2=c(1000,4000),
Sales_1=c(100,80),Sales_2=c(50,200))
df1 <- df %>%
mutate(Sales__1 = Revenue_1*Sales_1,
Sales__2 = Revenue_2*Sales_2) %>%
format(scientific=F)
Result
Revenue_1 Revenue_2 Sales_1 Sales_2 Sales__1 Sales__2
1 2000 1000 100 50 200000 50000
2 2500 4000 80 200 200000 800000

Answer 5

The tidyverse solution:
library(tidyr)
library(dplyr)
dat <- structure(list(Revenue_1 = c(2000, 2500), Revenue_2 = c(1000,
4000), Sales_1 = c(100, 80), Sales_2 = c(50, 200)), row.names = c(NA,
-2L), class = "data.frame")
dat %>%
mutate(id = row_number()) %>%
pivot_longer(-id, names_to = c(".value", "category"), names_pattern = "^(.*)_(.)$") %>%
mutate(Revenue_Sales = Revenue * Sales) %>%
pivot_wider(names_from = category, values_from = c(Revenue, Sales, Revenue_Sales)) %>%
select(-id)
# A tibble: 2 x 6
Revenue_1 Revenue_2 Sales_1 Sales_2 Revenue_Sales_1 Revenue_Sales_2
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 2000 1000 100 50 200000 50000
2 2500 4000 80 200 200000 800000