How to normalize data series to start value = 0?

Question

I have a dataset similar to this:

library(ggplot2)
data(economics_long)
economics_long$date2 <- as.numeric(economics_long$date) + 915
ggplot(economics_long, aes(date2, value01, colour = variable)) +
       geom_line()

Which gives the following plot:

Now I would like to normalize it to the start value of the green line (or the mean), so all variables start at the same point of the Y axes. Similar to this:

Thanks for any help.

Answer 1

You could subtract the starting value of each vector depending on variable -value using by() .

library(ggplot2)
l <- by(economics_long, economics_long$variable, function(x) 
  within(x, varnorm <- value01 - value01[1]))
dat <- do.call(rbind, l)

ggplot(dat, aes(date2, value01.n, colour = variable)) +
  geom_line()

Answer 2

use group_by() and mutate() to shift each variable by its initial y-value.

library(tidyverse)
data(economics_long)

economics_long %>% 
  group_by(variable) %>% 
  mutate(value_shifted = value01 - value01[1]) %>% 
  ungroup() %>% 
  ggplot(aes(date2, value_shifted, colour = variable)) +
  geom_line()