compute percentage reduction based on a year

Question

I have this dataframe:

dat = read.delim(text = "LM, year
  85.20000, 2020
  56.70000, 2021
  49.00000, 2022
  71.00000, 2023
  33.00000, 2024
  96.50000, 2025
  26.30000, 2026
  21.30000, 2027", sep = ",", header = T)

I want to calculate the percent reduction change from a specific year, I wrote this function regarding that:

pChange = function( dat , startYear){

  lastYear = max(dat$year)
  
  res = dat %>% arrange(year) %>%  
    mutate(change = (LM[year == lastYear] - LM[year == startYear]) / LM[year == startYear])
  res

}

pChange( dat, startYear = 2022)

However, I get the same value for every year, and I think there is problem but I don't how to fix it. Any idea?

    LM year     change
1 85.2 2020 -0.5653061
2 56.7 2021 -0.5653061
3 49.0 2022 -0.5653061
4 71.0 2023 -0.5653061
5 33.0 2024 -0.5653061
6 96.5 2025 -0.5653061
7 26.3 2026 -0.5653061
8 21.3 2027 -0.5653061

Answer 1

If you want to calculate the percent change between startYear and every year, I believe you don't want to include lastYear.

pChange = function(dat, startYear){
  
  res = dat %>% 
    arrange(year) %>%  
    mutate(change = (LM - LM[year == startYear]) / LM[year == startYear])
  res
}

pChange(dat, startYear = 2022)

    LM year     change
1 85.2 2020  0.7387755
2 56.7 2021  0.1571429
3 49.0 2022  0.0000000
4 71.0 2023  0.4489796
5 33.0 2024 -0.3265306
6 96.5 2025  0.9693878
7 26.3 2026 -0.4632653
8 21.3 2027 -0.5653061

Answer 2

I suggest you first calculate all percent changes from the previous year and then you subset the one you are looking for:

library(tidyverse)

dat <- tribble(
  ~ LM, ~ year,
  85.20000, 2020,
  56.70000, 2021,
  49.00000, 2022,
  71.00000, 2023,
  33.00000, 2024,
  96.50000, 2025,
  26.30000, 2026,
  21.30000, 2027,
)

pchange <- function(dat, start_year) {
  output <- dat %>%
    mutate(change = (LM - lag(LM, order_by = year)) / LM) %>%
    filter(year == start_year)
  output$change
}

pchange(dat, 2025)