Apply a growth rate to a given initial value conditionally within a data frame

Apologies if its a basic question but I simply cannot figure this one out as a beginner - I tried to find similar problem but was unsucessful...

I have a following data frame in which for every product and for every region combination the growth rate changes and is fixed for 10 years (I left_join it from another data frame).

I only have sales for the current year and want to project 10 year using growth rate given trying to produce results stated in sales_expected.

Within my pipeline I tried to use accumulate from purr package:

mutate(sales_expected = accumulate(sales, fill_in))

In combination with fill_in function which looks up the next n.a within sales column and applies growth rate:

fill_in <- function(prev, new, growth) {
  growth_rate <- 0.1
  if_else(!is.na(new), new, prev * (1 + growth_rate))
}

This works but what I cannot figure out is how to change the growth_rate argument within this function conditionally i.e. it should be picked up from column growth instead of me typing it manually as this obviously does not work with new product & region combinations being introduced into the data frame for which growth rate changes.

thanks so much for help!

You could do it with group_by and cumprod. cumprod(growth+1) will do the compounding of the growth rates; I used lag to set the growth for the first year to 0 (to make it the same as the baseline).

dat |> 
  group_by(product, region) |> 
  mutate(sales_projected = sales[1] * cumprod(lag(growth, default = 0)+1))

# A tibble: 13 × 7
# Groups:   product, region [2]
   product  region   year growth sales sales_expected sales_projected
   <chr>    <chr>   <int>  <dbl> <int>          <int>           <dbl>
 1 productA regionA  2023   0.1    100            100            100 
 2 productA regionA  2024   0.1     NA            110            110 
 3 productA regionA  2025   0.1     NA            121            121 
 4 productA regionA  2026   0.1     NA            133            133.
 5 productA regionA  2027   0.1     NA            146            146.
 6 productA regionA  2028   0.1     NA            161            161.
 7 productA regionA  2029   0.1     NA            177            177.
 8 productA regionA  2030   0.1     NA            195            195.
 9 productA regionA  2031   0.1     NA            214            214.
10 productA regionA  2032   0.1     NA            236            236.
11 productA regionB  2023   0.06   100            100            100 
12 productA regionB  2024   0.06    NA            106            106 
13 productA regionB  2025   0.06    NA            112            112.

If you want to use accumulate functions I think you need to use accumulate2() like this:

dat |> 
  group_by(product, region) |> 
  mutate(sales_expected = accumulate2(sales, growth[-1], \(x,y,g) x*(1+g)))

# A tibble: 13 × 7
# Groups:   product, region [2]
   product  region   year growth sales sales_expected sales_projected
   <chr>    <chr>   <int>  <dbl> <int>          <int>           <dbl>
 1 productA regionA  2023   0.1    100            100            100 
 2 productA regionA  2024   0.1     NA            110            110 
 3 productA regionA  2025   0.1     NA            121            121 
 4 productA regionA  2026   0.1     NA            133            133.
 5 productA regionA  2027   0.1     NA            146            146.
 6 productA regionA  2028   0.1     NA            161            161.
 7 productA regionA  2029   0.1     NA            177            177.
 8 productA regionA  2030   0.1     NA            195            195.
 9 productA regionA  2031   0.1     NA            214            214.
10 productA regionA  2032   0.1     NA            236            236.
11 productA regionB  2023   0.06   100            100            100 
12 productA regionB  2024   0.06    NA            106            106 
13 productA regionB  2025   0.06    NA            112            112.