Having trouble calculating something, am I missing a function ?

[chefsanglier]

Hello guys, I'm having trouble by doing something.
Imagine I got 4 columns (col1, col2,col3,col4)
I'm bad at explanation so a screen will be easier :
I would like a 5th column, who gives me, as an example here, 1000(col4) , split by every A (col2), and the split depend on the col3 (250,144,18) as proportion of 1000.
So col5 would be something like 600, 346,44 for the first three rows

forgot to add that col2 A will always correspond with col4 1000
every time there is an "A" at col2, col4 will always be "1000"
I'm having trouble searching on this forum because I don't know how to write my problem.
Thanks you !

[nirgrahamuk]

(df1 <- data.frame(
  col2 = c(rep("a", 3), rep("b", 3)),
  col3 = c(250, 144, 18, 1225, 4200, 1777),
  col4 = c(rep(1000, 3), rep(2000, 3))
))

library(tidyverse)

(df2 <- group_by(
  df1,
  col2
) %>% summarise(col3_sum = 
                  sum(col3,
                      na.rm = TRUE)) %>%
    ungroup())

(df3 <- left_join(df1, 
                  df2,
                  by = "col2") %>%
  mutate(col4x = 
           (col4 * col3) /
           col3_sum))