Monday, 21 May 2012

Alternate way of plotting means and errors

Last month, I wrote a post discussing dynamite plots, noting that they're not considered to be especially good at presenting information. I got a little bit of flak for it, from people for and against dynamite plots. This post shows a different method of showing a point and an error bar. If you're going to do it, why not make it look like something cool. A TIE fighter maybe?
Values <- c(1,2,5,4)
Errors <- c(0.25, 0.5, 0.33, 0.12)
dummyNames <- paste("Trial", 1:4)

TIEplot(1:4, Values, Errors, names = dummyNames, ylim = c(0,6))

How much cooler is that?
TIE fighter plot

The code for TIEplot() is available on gitHub. Remember to acknowledge me when you use in all your papers/books/theses.

7 comments:

Unknown said...

May I strongly urge that you give thought to violin plots - the modern take on the old boxplot, visually very appealing and readily explained to novice audiences. See, for example, R packages caroline, violinmplot and vioplot, as well as Stata's most recent version of this graphic.

D L McArthur said...

..

D L McArthur said...

Caroline's violins allows violins by groups, offset, overprinted and colorized, as well as extended statistics and other good things

Inkhorn said...

Hahahahahahahah

Inkhorn said...

Hahahahahahahah

a Tom said...

Brilliant! I'm looking forward to other SW icons in this series of intergalactic graphics.

Chris Beeley said...

Awesome! I'm having a terrible week and this cheered me right up. I'll be sure to let you know (and cite you!) when I use them in my career-defining paper in Science.