R: how to layer two facet ggplot from two different df with dates as the x axis?

Question

I would like to combine these two ggplots: origin from different data frames with the same X axis (floor_dates) and same facet_wrap. The two plots:

ggplot(averaged_activity,aes(x=date(DATETIME),y=PERCENTAGE), group = POPULATION) +
    geom_line(stat = "summary", fun = "mean")+ geom_point(stat = "summary", fun = "mean")+
    labs(y="Activity (%)", x = "Date")+
    scale_x_date(date_labels="%d/%m/%y",date_breaks  ="1 week")+
    scale_y_continuous()+
    facet_wrap(~POPULATION, ncol=1)

looks like: and

ggplot(info_table, aes(x=floor_date(DATE_OF_METAMORPHOSIS, unit = "week"), group= POPULATION))+
  geom_histogram()+facet_wrap(~POPULATION, ncol=1)

looks like: I would like the two plots to be combined on the same graph.

I've tried to add the histogram to the first ggplot with geom_histogram(info_table, aes(x=floor_date(DATE_OF_METAMORPHOSIS, unit = "week"), group= POPULATION)), I also tried to create the ggplot with no defined data frame and define the data frames inside the geoms. but every way I tried I got a different error, all about r having trouble finding variables, like the "POPULATION" variable in the facet or the group commend or the "PERCENTAGE" variable in the y axis.

tnx!

I recommend trying the patchwork package. https://patchwork.data-imaginist.com/ Within ggplot2 itself, it's difficult or impossible to make plots with separate geoms on separate facets. — Arthur, Sep 14 '22 at 17:02
It _is_ possible to do this, but very difficult without reproducible data to show you how. — Allan Cameron, Sep 14 '22 at 17:37
I dont mean separate geoms on separate facet, but on same 3 geoms (point, line and histogram) on the same facets, when the histogram originated from another data frame. — Maya Eldar, Sep 14 '22 at 18:15

score 1 · Accepted Answer · answered Sep 15 '22 at 06:52

Here is an option of pulling the two graph together. (As per comments, it would be best to include reproducible example in your questions)

Dummies Data

set.seed(12)

df1 = data.frame(surname = rep(c("Zaura", "Garin", "Roberts"), 20),
                 DATETIME = rep(seq(as.Date('2022-01-01'), as.Date('2022-10-01'), by = "2 weeks"), 3), 
                 percentage = sample(seq(50, 300, 50), 60, prob =c(1000, 100, 50, 10, 5, 1), replace = T))
  
df2 = data.frame(surname = sample(c("Zaura", "Garin", "Roberts"), 50, replace =TRUE),
                 DATE_OF = sample(seq(as.Date('2022-07-01'), as.Date('2022-09-01'), by = "2 weeks"), 50, prob = c(10,100,1,200,20), replace = TRUE))

Individuals plots

p1 = ggplot(df1, aes(x=DATETIME, y=percentage))+
  geom_line(stat = "summary", fun = "mean")+
  geom_point(stat = "summary", fun = "mean")+
  facet_wrap(~surname, ncol=1)+
  labs(subtitle = "Plot 1")

p2 = ggplot(df2, aes(x=floor_date(DATE_OF, unit= "week")))+
  geom_histogram(aes(y = stat(count)))+
  facet_wrap(~surname, ncol =1)+
  labs(subtitle = "Plot 2")

library(ggpubr)
ggarrange(p1,p2)

Merging the plots together

One option to merge the plot is to call for data= and aes() in each geom_. Plus one issue here is that the histogram count is around 10 max whereas the lines could be in the hundreds. The trick is to multiply the y = stat(count) here by 10 to have them overlay the points. Then in the scale_y_continuous() you can call for the second axis and do the inverse transformation. In our case dividing by 10 sec.axis = sec_axis(~ . / 10, name = "Count")

ggplot()+
  geom_histogram(data=df2, aes(x=floor_date(DATE_OF, unit= "week"), y = stat(count)*10), binwidth = 10)+
  geom_line(data=df1, aes(x=DATETIME, y=percentage), stat = "summary", fun = "mean")+
  geom_point(data=df1, aes(x=DATETIME, y=percentage), stat = "summary", fun = "mean")+
  scale_y_continuous("Percentage (%)", sec.axis = sec_axis(~ . / 10, name = "Count"))+
  facet_wrap(~surname, ncol=1)+
  labs(x="Date")

R: how to layer two facet ggplot from two different df with dates as the x axis?

1 Answers1

Dummies Data

Individuals plots

Merging the plots together