How to change the order of a bar plot(two categorical variables)

Question

So I got stuck with this problem for a while and can't solve it even after extensive research and experimentation, please help me out here.

I was trying to plot the relationship between education level and general health, here is my code.

p <- ggplot(educa_genhlth, aes(x = educa, fill = genhlth)) +
  geom_bar(position = "fill")
q <- p +
  aes(stringr::str_wrap(educa, 10)) + 
  labs(title = "general health vs education background") +
  xlab(NULL)
r <- q+
  scale_fill_discrete(name="general health")
r

Note I wrote line aes(string::str_wrap(Educa, 10)) because the labels of x variable were too long and clogged over one another and makes it hard to read. I searched this function at the suggestion of another post on this website.

But, a new problem is that the bar doesn't follow a logical order, i.e. say from " Never attended school " to "College 4 years ...". It was organized, I assume, alphabetically. So I did some research and realized that I have to give order to the factor variable educa_health$educa. So I added another line of code

educa_genhlth$educa <- factor(educa_genhlth$educa,
                              ordered = TRUE,
                              c("Never attended school or only kindergarten",
                                "Grades 1 through 8 (Elementary)",
                                "Grades 9 though 11 (Some high school)",
                                "Grade 12 or GED (High school graduate)",
                                "College 1 year to 3 years (Some college or technical school)",
                                "College 4 years or more (College graduate)"))

p <- ggplot(educa_genhlth, aes(x = educa, fill = genhlth)) +
  geom_bar(position = "fill")
q <- p +
  aes(stringr::str_wrap(educa, 10)) +
  labs(title = "general health vs education background") +
   xlab(NULL)
r <- q +
  scale_fill_discrete(name = "general health")
r

But it turned out it didn't change anything.

However, if I keep the line that give order to factor educa_health$educa but delete the part about wrap the string in line 4, I could however get the reorganized plot I want.(note that in order to see it more clearly I flip the plot horizontally by adding coord_flip())

educa_genhlth$educa <- factor(educa_genhlth$educa,
                              ordered = TRUE,
                              c("Never attended school or only kindergarten", 
                                "Grades 1 through 8 (Elementary)",
                                "Grades 9 though 11 (Some high school)",
                                "Grade 12 or GED (High school graduate)",
                                "College 1 year to 3 years (Some college or technical school)",
                                "College 4 years or more (College graduate)"))

p <- ggplot(educa_genhlth, aes(x = educa, fill = genhlth)) +
  geom_bar(position = "fill")
q <- p +
  labs(title = "general health vs education background") +
  xlab(NULL)
r <- q +
  scale_fill_discrete(name = "general health") +
  coord_flip()
r

I have absolutely at my wits' end. what I want is to keep the plot vertical, keep the labels readable and with a logical order I assigned. I really really appreciate if someone can tell me how to do it and why my original approach was ineffective.

Here is a small sample of my dataset:

structure(list(educa = structure(c(6L, 5L, 6L, 4L, 6L, 6L), .Label = c("Never attended school or only kindergarten", 
"Grades 1 through 8 (Elementary)", "Grades 9 though 11 (Some high school)", 
"Grade 12 or GED (High school graduate)", "College 1 year to 3 years (Some college or technical school)", 
"College 4 years or more (College graduate)"), class = "factor"), 
    genhlth = structure(c(4L, 3L, 3L, 2L, 3L, 2L), .Label = c("Excellent", 
    "Very good", "Good", "Fair", "Poor"), class = "factor")), row.names = c(NA, 
6L), class = "data.frame")

To keep the factor ordering, it needs to be the last step. When you took your ordered factor and then fed it into `str_wrap`, that converted it from a factor back into a character variable. Convert to a wrapped string and then to an ordered factor. — Jon Spring, Feb 03 '19 at 19:24

score 1 · Accepted Answer · answered Feb 04 '19 at 04:09

Assuming you've already ordered your educa in the desired order, you can use fct_relabel from the forcats package together with str_wrap, to change the factor labels in one step without converting it from character to factor again:

ggplot(educa_genhlth, 
       aes(x = forcats::fct_relabel(educa,
                                    stringr::str_wrap,
                                    width = 10), 
           fill = genhlth)) +
  geom_bar(position = "fill") +
  labs(title = "general health vs education background") +
  xlab(NULL) +
  scale_fill_discrete(name = "general health")

This approach also keeps the educa_genhlth$educa in the data frame in the original form, leaving you the flexibility to wrap it to other lengths in other plots.

score 0 · Answer 2 · answered Feb 03 '19 at 21:05

The use of str_wrap reorder your factors. So you need first to wrap, and then reorder your factors:

educa_genhlth$educa <- stringr::str_wrap(educa_genhlth$educa,10)
educa_genhlth$educa <-factor(educa_genhlth$educa,ordered=TRUE,
                             stringr::str_wrap(c("Never attended school or only kindergarten",
                               "Grades 1 through 8 (Elementary)",
                               "Grades 9 though 11 (Some high school)",
                               "Grade 12 or GED (High school graduate)",
                               "College 1 year to 3 years (Some college or technical school)",
                               "College 4 years or more (College graduate)"),10))




p<-ggplot(educa_genhlth,aes(x=educa,fill=genhlth))+geom_bar(position="fill")
q<-p+aes(educa)+labs(title="general health vs education background")+xlab(NULL)
r<-q+scale_fill_discrete(name="general health")
r

How to change the order of a bar plot(two categorical variables)

2 Answers2