Hello,
is there anyone who can give me some advice on creating histogram with split axis in R?
Thanks in advance
That wouldn't make sense, imho.
If you anyway want to do such nonsensible things, you may have a look at the plotrix package.
For back-to-back histograms, see https://www.rdocumentation.org/packages/Hmisc/versions/4.1-1/topics/histbackback, and I am sure these are in other packages too. Often for procedures like this it is easier to write the code in one of the main R graphics packages since these specialized functions can constrain what you can do (and sometimes don't let you pass all the parameters that you want to, to the main plotting functions which they usually call). Of course this takes time, so if you are doing this for EDA you may not want to bother.
It would useful to know what aspects of the data that you are hoping to examine. There are lots of different reasons for plotting data, and some procedures will be better suited for some of these (and some won't be sensible). For example, if wanting to compare the two groups, plotting the empirical cumulative distributions (ecdf in R) might make more sense or if to see how they each group compares with some theoretical distribution, a quantile-quantile plot could be useful. I do think for some purposes the readers do want to see both distributions and things like a series of boxplots don't show the aspects you wish to convey.
Not sure if this is what you mean, but with the ggplot graphics engine, facet_grid(~Group) splits a graph by a factor Group.
I am not sure what is your means, but you can remove axis of hist with axes=FALSE and you can add you wish axis.
For example:
x
thank you for your comments, I have solved it by saving the data as frequences and then ploted them as split barplot
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Software
- R