ggplot2 3D Bar Plot

Question

I Know this sounds basic, but have a been searching for literally more than an hour now without success. I'm simply trying to plot a 3D bar plot in 'R' using the 'ggplot2' package. My dataframe looks something like this:

 x   y     z
t1   5   high
t1   2   low
t1   4   med
t2   8   high
t2   1   low
t2   3   med
t3  50   high
t3  12   med
t3  35   low

and I want to plot something like this on it:

Any help is more than appreciated!!

Solution

As mentioned in comments, 3D plots usually aren't a good choice (when other options are available) since they tend to give a distorted/obscured view of data.

That said, here's how you can plot your data as desired with latticeExtra:

d <- read.table(text=' x   y     z
t1   5   high
t1   2   low
t1   4   med
t2   8   high
t2   1   low
t2   3   med
t3  50   high
t3  12   med
t3  35   low', header=TRUE)

library(latticeExtra)

cloud(y~x+z, d, panel.3d.cloud=panel.3dbars, col.facet='grey', 
      xbase=0.4, ybase=0.4, scales=list(arrows=FALSE, col=1), 
      par.settings = list(axis.line = list(col = "transparent")))

Comments

[IRTFM]

You should realize that R aficionados think of Excel as a rich source of examples of how NOT to do things (properly) and that 3D barplots are one of the most deprecated strategies for communication of accurate quantitative information.

[Tavi]

@BondedDust ha! yes, i do realise that. probably the reason why i cannot find any examples at all. but i still want to try, and any pointers from you will help… i just need to know how to add a third axis to my geom_bar()

[Docconcoct]

This may be an avenue worth investigating: stackoverflow.com/questions/23261760/…

[Tavi]

@Docconcoct thanks, but all the responses are for 2d plots!!

[Docconcoct]

True. Just thought it was worth consideration.

[IRTFM]

From the author of rgl: stat.ethz.ch/pipermail/r-help/2007-September/141578.html and from @hadley: vita.had.co.nz/papers/prodplots.pdf

[eipi10]

I don't think you can make a 3D barplot with ggplot2. Like exceeding the speed of light in our universe, there are just some things that are forbidden by the fundamental laws of the Hadleyverse. In any case, a line plot grouped by color will be much more informative. For example, with the data you listed above try this (where I've named the data frame dat): ggplot(dat, aes(x, y, group=z, colour=z)) + geom_line() + geom_point().

[tonytonov]

@eipi10 Could not agree more. This comment of yours is actually worth posting as a decent answer IMO.

[jbaums]

Does it have to be ggplot? (If not, take a peek at the examples in ?latticeExtra::panel.3dbars.

[Tavi]

@jbaums thank u J I'll check it now. No it doesn't have to be ggplot. Anything I can do in R would be just fine :)

[Tavi]

@eipi10 thank you I'll try that in a second and post it as answer possibly;)

[Tavi]

@tonytonov good to know I haven't tried it just yet as I've been away but will do in a second!!

[jbaums]

@maryam - just noticed the levels of z are ordered incorrectly. You can correct that by explicitly stating the level order when creating the factor.

[Tavi]

oh yeah, it goes med -> low -> high instead of high -> med -> low thanks for the correction :)

[DavideChicco.it]

What does `y~x+z' mean?

[jbaums]

@DavideChicco.it - it says to plot y against x and z (where these are elements of data d).

[DavideChicco.it]

@jbaums Thanks, I should have read the documentation. Another little help: how could I increase the dimensions of the sides of the parallelepipedon? Thanks!

[Matthew Son]

The code raises error : Error Using Packet 1 Non-numeric argument to binary operator