I am trying this very simple code:

mtcars %>% 
  mutate(Bla = cyl+mpg) %>% 
  ggplot(. %>% 
           filter(am == 1), aes(x = mpg, y = cyl)) +
  geom_point()

It doesn’t work and gives out an error You've supplied a <fseq> object. However, the following works:

ggplot(mtcars %>% 
         mutate(Bla = cyl+mpg) %>% 
         filter(am == 1), aes(x = mpg, y = cyl)) +
  geom_point()

Therefore, it seems that it’s not possible to pass the dot . argument to ggplot and then follow it along with a pipeline.

I swear that I used to be able to do this, therefore my questions are:

1. Was this changed at some point in ggplot and am I just doing the old thing; or was this never possible and I am doing it completely wrong?
2. If this was indeed changed, then what is now the preferred and suggested ways to pass these dot arguments?