Two plots overlaid (ggplot)

Question

Can anyone help me in my struggle to combine these two graphs into one? Ideally the coloured plot should be overlaid by the step plot, so that counts of "ox_0" are shown (as histogram or steps) on the coloured graph.

I have PLOT 1:

ggplot(subset(stw, !is.na(will)), aes(ox_0, fill=will)) +
    geom_bar(position="fill") +
    ggtitle("Willingness to repeat surgery per Oxford Knee Score at baseline") +
    xlab("Preoperative Oxford Knee Score")+ ylab("")+
    scale_y_continuous("", labels=c("0 %", "25%", "50%", "75%", "100%"))+
    theme(
    axis.text.y = element_text(size=12, color=1),
    axis.title.x = element_text(size=15, color=1),
    axis.text.x = element_text(size=12, color=1),
    axis.title.y = element_text(size=15, color=1),
    plot.title = element_text(hjust=0.5, size=16),
    strip.text.x = element_text(size=12),
    legend.text = element_text(size=13),
    legend.title = element_text(size=15))+
    scale_fill_manual("Would you repeat surgery?", values=col.traf, aesthetics = "fill")

and PLOT 2:

ggplot(subset(stw, !is.na(will)), aes(ox_0))+
    stat_bin(geom="step", binwidth = 1)

It's easier to help you if you include a simple [reproducible example](https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example) with sample input and desired output that can be used to test and verify possible solutions. — MrFlick, Feb 28 '23 at 23:23
In general you could add the stat_bin layer to your first plot, i.e try with `+ stat_bin(data = subset(stw, !is.na(will)), aes(ox_0), geom="step", binwidth = 1, inherit.aes = FALSE)`. However, your probably have to rescale the count for the stat_bin because of the different scales. For more help see the comment by MrFlick. — stefan, Feb 28 '23 at 23:48
Thanks to you both for helping me. I used this other solution below, because it seemed easier for me, and that worked. — Anne MP, Mar 01 '23 at 08:15

Jon Spring · Accepted Answer · 2023-03-01T01:50:00.053

Here's an example with a dataset from dplyr. In this case I put the fill aesthetic into only the geom_bar layer, and use y = after_stat(count)/1000 to bring the range of the 2nd layer (max value 684) into the range 0:1 that the "fill" bars will necessarily cover. Here I add a secondary axis label to show how the step line should be interpreted.

ggplot(dplyr::storms, aes(day)) +
  geom_bar(aes(fill = status), position = "fill") +
  stat_bin(geom = "step", binwidth = 1, aes(y = after_stat(count)/1000)) +
  scale_y_continuous(sec.axis = sec_axis(~ . *1000))

Two plots overlaid (ggplot)

1 Answers1