I am trying to create a grouped and stacked bar chart that shows proprtions by a group and then further differentiated by another group. I would like my chart to look something like this But all of my code shows my chart like this This feels like a very silly problem because proportions are so simple but I really can not make it grouped by main groups and then by two different groups and then stacked by subgroups.

My data is here

I tried:

figure4 <- figure1_edit %>% 
  summarise(N = n(), .by = c(group, carbon_chain, `Functional Group`)) |> 
  mutate(
    prop = (N / sum(N))*100,
    x = paste0(group, "n", carbon_chain),
    .by = c(group)
    ) %>%
  ggplot(aes(y = carbon_chain, x = prop, fill = `Functional Group`)) + 
  geom_bar(stat = "identity", position = "fill") + 
  scale_x_continuous(labels = scales::percent_format(scale = 1)) + 
  facet_grid(~ group) + 
  scale_fill_brewer(palette = "Set3") + 
  theme_bw()
  
figure4

But this gave me the figure that is not correctly differentiated. So then I tried to create sub groups by the original groups (so in this example I just tried the ‘rain’ group), calculate the proportions needed, but this still gave me my incorrect chart.

I would like it to be grouped by ‘group’ (so essentially 3 different bar charts) and then within each group, grouped by the chain length (short vs long) and the % of short and long found per the ‘group’ of interest, and within each chain length further filled by the % of functional group per each chain length group.