I noticed when combining multiple ggplots, where the values in one are much larger than the other, because the y axis ticks are larger in the former (due to being bigger numbers), it makes the plot itself smaller. To illustrate:
df <- data.frame('name'=1:30,'v1'=rpois(30, 1),'v2'=rpois(30, 100))
p1 <- ggplot(data=df,aes(x=name,y=v1)) + geom_bar(stat = 'identity')
p2 <- ggplot(data=df,aes(x=name,y=v2)) + geom_bar(stat = 'identity')
combined <- ggarrange(p1, p2, ncol=1, nrow=2)
print(combined)

Note that the bottom plot is slightly smaller than the top plot, because the the numbers in the y-ticks are larger. Is there anyway to make it so that the plots themselves are exactly the same size? (That doesn't involve manually changing the tick sizes until the plots look approximately the same). I tried using ggplotGrob(), which was suggested elsewhere, but it didn't seem to help. Thanks.

