41

I am trying to make a y-axis title with both a special character and a superscript. I am able to do this, but I want the closing parentheses not to be superscripted. And that's what I'm having a problem with. I think its just a placing of my parenthesis, but I've tried (seemingly) everything. 

plot(WatexCl, ConcuM, col = as.numeric(1), pch = as.numeric(Depth), 
   xlab = expression(paste("Concentration Cl  ( ", mu, "moles/g dry wt)")), 
   ylab = expression(paste("Average Conc of S- on plates ( ", mu, "Moles/cm"^"2"),)), 
   data = plates)
zx8754
  • 52,746
  • 12
  • 114
  • 209
Valerie S
  • 917
  • 2
  • 7
  • 17

2 Answers2

73

The one thing that often users fail to grasp is that you invariably don't need to quote strings and paste them together when used in an expression for a plot label. It is usually simpler to use the layout tools directly (e.g. ~ and *). For example:

df <- data.frame(y = rnorm(100), x = rnorm(100))

plot(y ~ x, data = df,
     ylab = expression(Average ~ Conc ~ of ~ S- ~ on ~ plates ~ 
                       (mu ~ Moles ~ cm^{-2} ~ dry ~ wt)),
     xlab = expression(Concentration ~ Cl ~ (mu ~ moles ~ g^{-1} ~ dry ~ wt)))

Alternatively, you can include strings for longer sections of text; in this case it is arguably easier to do:

plot(y ~ x, data = df,
     ylab = expression("Average Conc of S- on plates" ~
                         (mu ~ moles ~ cm^{-2} ~ "dry wt")),
     xlab = expression("Concentration Cl" ~ (mu ~ moles ~ g^{-1} ~ "dry wt")))

but note there is no need to paste strings and other features here.

Both produce:

enter image description here

Note the issue plotmath has with the superscript 2. You may wish to add some extra space for the y-axis margin to accommodate that:

op <- par(mar = c(5,4.5,4,2) + 0.1)
plot(y ~ x, data = df,
     ylab = expression("Average Conc of S- on plates" ~
                          (mu ~ moles ~ cm^{-2} ~ "dry wt")),
     xlab = expression("Concentration Cl" ~ (mu ~ moles ~ g^{-1} ~ "dry wt")))
par(op)

producing

enter image description here

Gavin Simpson
  • 170,508
  • 25
  • 396
  • 453
7

This solves the problem of the super-scripted closing parenthesis: 

# reproducible data
plates <- data.frame(WatexCl = rnorm(100), ConcuM = rnorm(100), Depth = rnorm(100))

# alter the default plot margins so the 
# superscript in the y-axis label is completely displayed
par(mar=c(5,5,4,2))

# draw the plot
plot(WatexCl ~ ConcuM, data = plates,
     col = as.numeric(1), 
     pch = as.numeric(Depth), 
     xlab = bquote("Concentration Cl ("*mu~"moles/g dry wt)"), 
     ylab = bquote("Average Conc of S- on plates ("~mu~"Moles/cm"^"2"*")"))

enter image description here

Ben
  • 41,615
  • 18
  • 132
  • 227
  • 2
    `bquote` is a bit of overkill here. `expression` should work fine. `bquote` is most useful when one wants to replace an object in the expression with its value. – Gavin Simpson Apr 01 '13 at 01:38
  • @GavinSimpson, thanks, that's good to know! Looks like `expression` also does a better job with spacing characters than `bquote`, I've got a few odd spaces near the parentheses on my plot... – Ben Apr 01 '13 at 01:53
  • 6
    That is because you use `~ mu ~` to separate the two quoted strings. If you set that to `* mu ~` you'll get the right spacing. `~` is a spacing operator, `*` juxtaposes the arguments (i.e. without a space). – Gavin Simpson Apr 01 '13 at 01:55