Stacked bar chart with percentage labels - r

Question

I created a stacked bar chart where the bars represent a percentage of the population. I would like to add labels to the 65+ category (or for all 3 categories if it is not possible to do it just for 1 category) showing the % value for each year. If I add geom_text(label = datm$value), the bars become extremely small because the labels represent absolute values instead of percentages. This is my code:
dat <- read.table(text = "2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
0-20 24.0 23.9 23.7 23.5 23.3 23.1 22.9 22.7 22.5 22.3 22.2
20-65 61.3 61.2 61.0 60.9 60.5 60.1 59.8 59.6 59.3 59.1 59.0
65+ 14.8 15.0 15.3 15.6 16.2 16.8 17.4 17.7 18.2 18.5 18.8", sep = " ", header = TRUE)
library(reshape)
datm <- melt(cbind(dat, ind = rownames(dat)), id.vars = c('ind'))
library(scales)
library(ggplot2)
ggplot(datm,aes(x = variable, y = value, fill = ind)) +
geom_bar(position = "fill",stat = "identity") +
scale_x_discrete(labels = c('2008', '2009', '2010', '2011', '2012', '2013',
'2014', '2015', '2016', '2017', '2018')) +
scale_y_continuous(labels = percent_format()) +
xlab('Year') +
ylab('% of population') +
ggtitle('Demographic trend in the Netherlands') +
scale_fill_manual(values = c("green", "blue", "darkgray"))

Answer 1

You can try this. Explanations in comments below:
library(dplyr)
# calculate percentage within each year
datm2 <- datm %>%
group_by(variable) %>%
mutate(p = value / sum(value)) %>%
ungroup()
> head(datm2)
# A tibble: 6 x 4
ind variable value p
<fct> <fct> <dbl> <dbl>
1 0-20 X2008 24 0.240
2 20-65 X2008 61.3 0.612
3 65+ X2008 14.8 0.148
4 0-20 X2009 23.9 0.239
5 20-65 X2009 61.2 0.611
6 65+ X2009 15 0.150
ggplot(datm2, aes(x = variable, y = value, fill = ind)) +
geom_col(position = "fill") + # geom_col is equivalent to geom_bar(stat = "identity")
geom_text(aes(label = scales::percent(p), # add layer for percentage values
alpha = ifelse(ind == "65+", 1, 0)), # only visible for 65+ category
position = position_fill(vjust = 0.5)) + # follow barplot's position
scale_x_discrete(labels = c('2008', '2009', '2010', '2011', '2012', '2013',
'2014', '2015', '2016', '2017', '2018')) +
scale_y_continuous(labels = percent_format()) +
scale_alpha_identity() +
xlab('Year') +
ylab('% of population') +
ggtitle('Demographic trend in the Netherlands') +
scale_fill_manual(values = c("green", "blue", "darkgray"))