Panels are different sizes when converting ggplot plot with facets

When converting a ggplot plot with facets to a plotly object the panels are different sizes. Margins are not subtracted from top and bottom rows so these panels end up being taller. This makes it hard to compare among panels. Margins for top and bottom of panels are equal so longer x-axis labels result in large margins on tops of panels with large spacing between panels and a low data-ink ratio.

library(dplyr)
library(ggplot2)
library(plotly)
library(reshape2)

p <- ggplot(data = tips,
            aes(x = sex, y = total_bill)) +
  geom_jitter(width = 0.1, height = 0) +
  facet_wrap(day ~ time, scales = "free_x", ncol = 2) +
  theme(axis.text.x = element_text(angle = 45),
        strip.text.x = element_text(size = 8, 
                                    margin = margin(t = 1, r = 0, b = 1, l = 0, unit = "pt")))

q <- ggplotly(p) %>% 
  plotly::layout(margin = list(l = 80, r = 0, t = 100, b = 100, pad = 0))

d <- rbind(q$x$layout$yaxis4$domain, q$x$layout$yaxis3$domain, q$x$layout$yaxis2$domain, q$x$layout$yaxis$domain)
df <- data_frame(lb = d[,1], ub = d[,2]) %>% 
  mutate(range = ub - lb,
         panelSpacing = lb - lag(ub))

Reproduisez le problème avec l’exemple ggplotly() fourni et examinez les valeurs du domaine de l’axe y collectées dans d. Commencez par retracer la manière dont les panneaux de facettes et les marges de mise en page sont convertis, puis vérifiez que les panneaux supérieur et inférieur ont la même taille et que l’espacement reflète les marges configurées.