Find the percentage change and grouping within a dataframe-CodePudding

This is more of an extension to a question from earlier I was trying to work out the percentage change of variables between years. I now wish to extend this by doing it so I can see the change in countries through the years. Some sample data and code included below.

This code works out % change used, but I need to add in counties to this. Think some sort of group_by() on the county names might work here, although can't think of ways to use it here

df |>
  mutate(pdiff = 100*(gdp_per_capita - lag(gdp_per_capita))/gdp_per_capita)

data sample

df
Country_Name year gdp_per_capita
Albania 2018       5287.664
Albania 2019       5396.216
Albania 2020       5332.160
Albania 2021       6494.386
Algeria 2018       4142.019
Algeria 2019       3989.668
Algeria 2020       3306.858
Algeria 2021       3765.035
Bosnia 2018        6070.353
Bosnia 2019        6119.762
Bosnia 2020        6082.367
Bosnia 2021        6916.438 
Croatia 2018       15227.560
Croatia 2019       15311.767
Croatia 2020       14132.487
Croatia 2021       17398.766

CodePudding user response：

It could be done by adding the group_by

library(dplyr)
df |> 
  group_by(Country_Name) |> 
  mutate(pdiff = 100*(gdp_per_capita - lag(gdp_per_capita))/gdp_per_capita) |>    
  ungroup()

CodePudding user response：

I tried diff instead of using lag, but basically you need to group by country and do the same calculations. I dropped arrange in there to be sure about the order of the data.

library(dplyr)

df %>% 
  group_by(Country_Name) %>% 
  arrange(Country_Name, year) %>% 
  mutate(pdiff = 100 * c(NA, diff(gdp_per_capita) / gdp_per_capita[-1]))
  # mutate(pdiff = 100 * c(NA, diff(gdp_per_capita) / gdp_per_capita[-n()]))
  ## better to divide by previous row, i.e. lag(gdp_per_capita)

#> # A tibble: 16 x 4
#> # Groups:   Country_Name [4]
#>    Country_Name  year gdp_per_capita   pdiff
#>    <chr>        <int>          <dbl>   <dbl>
#>  1 Albania       2018          5288.  NA    
#>  2 Albania       2019          5396.   2.01 
#>  3 Albania       2020          5332.  -1.20 
#>  4 Albania       2021          6494.  17.9  
#>  5 Algeria       2018          4142.  NA    
#>  6 Algeria       2019          3990.  -3.82 
#>  7 Algeria       2020          3307. -20.6  
#>  8 Algeria       2021          3765.  12.2  
#>  9 Bosnia        2018          6070.  NA    
#> 10 Bosnia        2019          6120.   0.807
#> 11 Bosnia        2020          6082.  -0.615
#> 12 Bosnia        2021          6916.  12.1  
#> 13 Croatia       2018         15228.  NA    
#> 14 Croatia       2019         15312.   0.550
#> 15 Croatia       2020         14132.  -8.34 
#> 16 Croatia       2021         17399.  18.8

Data:

df <- read.table(text = "Country_Name year gdp_per_capita
Albania 2018       5287.664
Albania 2019       5396.216
Albania 2020       5332.160
Albania 2021       6494.386
Algeria 2018       4142.019
Algeria 2019       3989.668
Algeria 2020       3306.858
Algeria 2021       3765.035
Bosnia 2018        6070.353
Bosnia 2019        6119.762
Bosnia 2020        6082.367
Bosnia 2021        6916.438 
Croatia 2018       15227.560
Croatia 2019       15311.767
Croatia 2020       14132.487
Croatia 2021       17398.766", header = T, stringsAsFactor = F)