Plot two variables using ggplot2 (or other)

Question

I have a dataset called fruits_by_number like this:

category   Number
Apple     4
Banana    5 
Orange    2
Pear      0

I want to make a graph that on the x axis has bars symbolizing the fruit category and on the y axis the count as a height.

I have tried this:

qplot(x = category , y = number, data = fruits_by_number, geom = "dot")

My first question is: How instead of "dot" I ask for a bar?

My second questions is: Even if I get a bar, since in the (actual) dataset the fruit names are about 30, underneath the x-axis legend I get a big mess.

I saw this:

qplot(data = movies_df,Runtime,fill = Genre,bins = 30)

On this website there is this image:

this image

which is something close to what I want to get.

I tried to substitute fill with dodge, but it did not work.

Could you help me on how to make a graph that the fruit categories will be arranged by color on the right (like in the image) and then I will have colorfur non-overlappping bars with height equal to the number?

PS1: If I follow this recommendation:

p <- ggplot(data = trans_units_by_subject, aes(x = category, y = sumUnits_t, fill = category))
p + geom_bar(stat = "identity")

I still get underneath the x-axis the descriptions like this: enter image description here

If you have 30 fruits, color will not work to distinguish them. Anything more than about 8 colors is very hard to tell apart. — Gregor Thomas
– Gregor Thomas, Commented Dec 12, 2017 at 19:28

Alfredo Ascanio · Accepted Answer · 2017-12-12 20:29:55Z

3

I would recommend you use ggplot() instead of qplot(). I believe it's more manageable and you can play better with the options of the plot. Try something like this:

p <- ggplot(data = fruits_by_number, aes(x = category, y = Number, fill = category))
p + geom_bar(stat = "identity")

This will produce a barplot in which the X-axis will be the fruit category, Y-axis will be the number and you will have a color fill for each fruit in the plot, including its legend. I have to say to you that it's kind of redundant to use a color filling in this kind of graphs (excluding any aesthetic function) because all the fruits will have labels in the X-axis.

Another thing, you use "dodge" when you have two factors and several levels in each of them. For example suposse that you have an extra factor "Age" in your data frame: Apple (New/Old), Pear (New/Old). So you can use something like this:

p <- ggplot(data = fruits_by_number, aes(x = category, y = Number, fill = Age))
p + geom_bar(stat = "identity", position = "dodge")

The plot produced will have only two colors for factor Age(New/Old), and the bars will be separated. In the example that you put, the default option is position = "stack", so you will have stacked bars.

If you still want to use colors and not having the x-axis labels, use something like this:

p <- ggplot(data = fruits_by_number, aes(x = category, y = Number, fill = category))
p <- p + geom_bar(stat = "identity") 
p <- p + theme(axis.text.x = element_blank())
p

And try different color palettes

Check this pages out:

http://ggplot2.tidyverse.org/reference/geom_bar.html http://ggplot2.tidyverse.org/reference/theme.html

answered Dec 12, 2017 at 20:29

Alfredo Ascanio

815 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

Iniciador Over a year ago

Thank you - You said "all the fruits will have labels in the X-axis" you are correct, but I am trying to avoid that. My categories are about 30 so on the x-axis I don't get something legible. So can I while using the legentd describing the color fill not have the category legend appearing on the bottom? I have updated my question with that image of the legend underneath x.

Iniciador Over a year ago

And the second recommendation worked great! - I just included the image in my PS1, just to show you why I did not want the x-axis labeled.

Alfredo Ascanio Over a year ago

Glad it was helpful! I hope you keep using ggplot2. For future reference, remember to check the links I put at the end of my answer. You can use the theme() to change many things, including the aspect of your axis labels, font, fontsize, etc. Try this instead of the element_blank(), and you can rotate your axis labels in z degrees: theme(axis.text.x = element_text(angle = z))

ssp3nc3r · Accepted Answer · 2017-12-12 22:31:40Z

2

I suggest turning your plot sideways to accommodate the 30 categories. Using your data above,

require(ggplot2)

ggplot(fruits_by_number) + 
  geom_bar(aes(x = category, 
               y = Number, 
               fill = category), 
           stat = "identity") + 
  coord_flip()

where the above fill adds colors to the bars (remove this to make them all the same color), and the data is,

fruits_by_number <- 
read.table(text='
category   Number
Apple     4
Banana    5 
Orange    2
Pear      0', 
header=T)

edited Dec 12, 2017 at 22:31

answered Dec 12, 2017 at 20:29

ssp3nc3r

3,8422 gold badges16 silver badges25 bronze badges

2 Comments

Iniciador Over a year ago

Very helpful thank you - the rotation really helped. I did not know it was even an option.

ssp3nc3r Over a year ago

Sure. You can pretty much plot whatever you want with ggplot2. Here's it's online documentation for the basics: ggplot2.tidyverse.org

Antonio · Accepted Answer · 2017-12-12 20:19:25Z

1

Here is a possible solution using a built-in dataset:

mtcars %>%
mutate(CARS = factor(1:32)) %>%
ggplot(aes(CARS,mpg, fill = CARS)) +
geom_bar(stat = "identity")

answered Dec 12, 2017 at 20:19

Antonio

3271 silver badge11 bronze badges

1 Comment

Mako212 Over a year ago

It's probably a better practice to map fill to mpg rather than CARS because coloring by a generated sequential ID provides 0 additional information.

Collectives™ on Stack Overflow

Plot two variables using ggplot2 (or other)

3 Answers 3

3 Comments

2 Comments

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

3 Comments

2 Comments

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related