As Stephen points out, there is not enough space to display all the x-axis values. So rotating the values on the x-axis would help.

You can easily do this using the par() function while making a box plot of yvalues~xvalues as follows:

par(las=2)

boxplot(yvalues~xvalues)

You may also want to reduce the text size of the xvalues. For this the easiest way is to save the boxplot to an object (e.g. bp) and suppress the printing of the xvalues. Then use the axis() function to specify the xvalues and reduce the font size (using cex.axis). Change the cex.axis values to find the text size that suits you.

par(las=2)

bp

R drops some labels because there is not enough space that all labels can be shown without overlapping each other.

Solution 1:

use a smaller font size for the axis labels.

you can use the argument cex.axis=x (with 0

Thank you all specially Jochen Wilhelm. I am beginner in R and your answers help me very much. 

Dear Jochen Wilhelm 

Could you please send me R code for second plot?

Thank you

I have not saved it, but that's simple.

You can delete the line darwing the "level 3" x-axis. The colors are set by specifying the argument col in boxplot as a vector of 16 times "orange" and 16 times "skyblue", created with

col = rep(c("oragnge","skyblue"), each=16)

The legend was added with the function legend(), like

legend("bottomleft", fill=c("orange","skyblue"), labels=lvl3, bty="n")

Good luck :)

Thank you Jochen Wilhelm

I would love to create a similar boxplot but I am having problems in adjusting the R-code of Jochen to my data.

In my data set (ac.hu.green) have one numerical response variables (oral_disk) and 3 categorical (species, location and depth). Do you know perhaps what code I could use?

I tried doing a boxplot with this function :boxplot(oral_disk~species*location*depth, data=ac.hu.green, notch=TRUE,

col=(c("darkgreen","green")),

main="G-values of Oral disk HU_AC", xlab="Location and Depth")

But its very confusing and does not lable all the x values

Thank you in advance for your help

Try to set par(las=2), so the labels are drawn orthogonal to the axes.

[Edit: chnaged las=1 to las=2]

Do you mean :

boxplot(oral_disk~species*location*depth,par(las=1), data=ac.hu.green, notch=FALSE,

col=(c("darkgreen","green")),

main="G-values of Oral disk HU_AC", xlab="Location and Depth")

The funtion changes the boxplot a bit but the x-axis are still the same as the previous plot

Not quite. Either you set the graphical parameter "las" first in a call to par() and then call boxplot():

par(las=1)

boxplot(oral_disk~species*location*depth,data=ac.hu.green, notch=FALSE, col=(c("darkgreen","green"))

or you set it directly in the call to boxplot (not as a call to par(); the argument las is passed over to par(), what is called inside boxplot()):

boxplot(oral_disk~species*location*depth,data=ac.hu.green, notch=FALSE, col=(c("darkgreen","green"), las=1)

The first version is easier to read and to understand, the second is a bit shorter to write.

Thank you for your answer. Interesting though, it does not seem to change anything.

Sorry, you are right. You'll have to set las = 2 or 3, not 1.

See attached figure. The plot is produced with

par(mfcol=c(1,4))

for (i in 0:3) boxplot(InsectSprays, las=i, main=paste("las =",i))

Thank you very much for your help. Now i am getting all the values on the x axis (which is much better than before). Any chance you know how to put them in subgroups as you did in the very fist picture? In this way the lable would be a bit more clear, avoiding repetition of some values. It is also putting the picture to the very far right of the image.

Split the data by one variable, e.g. by "depth", then plot by the remaining variables:

splitted