Why isn't my barplot rearranging properly when faceting with ggplot? - r

Question

So I have made this barplot with this code, bars organised in descending order, great!
na.omit(insect_tally_native_ranges)%>%
group_by(native_ranges)%>%
dplyr::summarise(freq=sum(n))%>%
ggplot(aes(x=reorder(native_ranges,freq),y=freq))+
geom_col(color="#CD4F39",fill="#CD4F39",alpha=0.8)+
coord_flip()+
labs(x="Native ranges",
y="Number of invasive insect arrivals",
title="Species by native ranges")+
theme_minimal()
And now I wanted to do the same but faceting by a variable called Period, here's the code:
ggplot(native_freq_period,
aes(y=reorder(native_ranges,freq),x=freq))+
geom_barh(stat= "identity",
color="#CD4F39",
fill="#CD4F39",
alpha=0.8)+
labs(x="Native ranges",
y="Number of invasive insect arrivals",
title="Species by native ranges")+
theme_minimal()+
facet_wrap(~Period)
But the plot came out like this:
Which is pretty annoying because it is the same code as above and the levels for the variable native_ranges should be organised again. But instead it gives me this lumpy order that isn't even the alphabetic order. So the reorder part is reordering but not by freq! Don't understand.
Here is the data:
structure(list(native_ranges = structure(c(6L, 10L, 11L, 7L,
3L, 5L, 1L, 1L, 8L, 6L, 3L, 5L, 2L, 4L, 5L, 7L, 7L, 7L, 8L, 9L,
11L), .Label = c("Afrotropic", "Afrotropic/Neotropic", "Australasia",
"Australasia/Neotropic", "Indomalaya", "Nearctic", "Neotropic",
"Neotropic/Nearctic", "Neotropic/Nearctic/Australasia", "Palearctic",
"Palearctic/Indomalaya"), class = "factor"), Period = structure(c(4L,
4L, 4L, 4L, 4L, 4L, 3L, 4L, 4L, 3L, 3L, 3L, 4L, 4L, 2L, 1L, 2L,
3L, 2L, 4L, 3L), .Label = c("1896-1925", "1926-1955", "1956-1985",
"1986-2018"), class = "factor"), freq = c(21L, 13L, 12L, 11L,
10L, 10L, 4L, 4L, 4L, 3L, 2L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 1L,
1L, 1L)), row.names = c(NA, -21L), class = c("grouped_df", "tbl_df",
"tbl", "data.frame"), vars = "native_ranges", drop = TRUE, indices = list(
6:7, 12L, c(4L, 10L), 13L, c(5L, 11L, 14L), c(0L, 9L), c(3L,
15L, 16L, 17L), c(8L, 18L), 19L, 1L, c(2L, 20L)), group_sizes = c(2L,
1L, 2L, 1L, 3L, 2L, 4L, 2L, 1L, 1L, 2L), biggest_group_size = 4L, labels = structure(list(
native_ranges = structure(1:11, .Label = c("Afrotropic",
"Afrotropic/Neotropic", "Australasia", "Australasia/Neotropic",
"Indomalaya", "Nearctic", "Neotropic", "Neotropic/Nearctic",
"Neotropic/Nearctic/Australasia", "Palearctic", "Palearctic/Indomalaya"
), class = "factor")), row.names = c(NA, -11L), class = "data.frame", vars = "native_ranges", drop = TRUE))

Answer 1

You have to arrange the order of the variable first before plotting. Since you didn't provide any reproducible data I am using the following data
drugs <- data.frame(drug = c("a", "b", "c"), effect = c(4.2, 9.7, 6.1))
ggplot(drugs, aes(drug, effect)) +
geom_col()
Now to change the order of the variable use factor
drugs$drug <- factor(drugs$drug,levels = c("b","a","c")) #This is the order I want
ggplot(drugs, aes(drug, effect)) +
geom_col()
Here I provided the levels in factor manually. You can either provide them manually or sort the order of the variable first separately and provide. See below,
drugs$drug <- factor(drugs$drug,levels = drugs[order(drugs$effect),]$drug)
ggplot(drugs, aes(drug, effect)) +
geom_col()
This should work with facet_wrap as well.

Answer 2

OK, finally figured it out with help from the other answer. You need to create another column that summarizes the total frequency so you can then reorder by that column. There may be a more efficient way to do it, but I create a new summary data.frame and then join it back to the original and then reorder based on the new column.
summary_data <- data %>%
ungroup() %>%
group_by(native_ranges) %>%
summarize(total = sum(freq))
data <- data %>%
left_join(summary_data)
ggplot(data, aes(y = reorder(native_ranges, total),x = freq)) +
geom_barh(stat= "identity",
color="#CD4F39",
fill="#CD4F39",
alpha=0.8) +
labs(x="Native ranges",
y="Number of invasive insect arrivals",
title="Species by native ranges") +
theme_minimal()+
facet_wrap(~Period)