Skip to content

Typographical error in the documentation of function strwidth?

4 messages · ALBERTO VIEIRA FERREIRA MONTEIRO, Peter Dalgaard

#
There might be a minor typographical error in the documentation of the function strwidth, namely, from

?strwidth

we get:

"Note that the ?height? of a string is determined only by the number of linefeeds ("\n") it contains: it is the (number of linefeeds - 1) times the line spacing plus the height of "M" in the selected font. (...)"

I think the correct phrasing should be:

"Note that the ?height? of a string is determined only by the number of linefeeds ("\n") it contains: it is the (number of linefeeds) times the line spacing plus the (number of linefeeds + 1) times the height of "M" in the selected font. (...)"

Alberto Monteiro
#
I don't think so. I'll give you that it should either be the (number of lines - 1)*spacing or (number of linefeeds)*spacing, but it's correct to count the height of "M" only on the top line.

-pd

It is correct as written

  
    
#
Peter Dalgaard wrote:

            
Ah, ok, I see now. The "line spacing" is not the line spacing, but the inter-line spacing plus
the height of "M":

"Note that the ?height? of a string is determined only by the number of linefeeds ("\n") it contains: it is the (number of linefeeds - 1) times the line spacing plus the height of "M" in the selected font. For an expression it is the height of the bounding box as computed by plotmath. Thus in both cases it is an estimate of how far above the final baseline the typeset object extends. (It may also extend below the baseline.) The inter-line spacing is controlled by cex, par("lheight") and the ?point size? (but not the actual font in use)"

It's still confusing... Specially since there's no way to get the inter-line spacing except by running

strheight("M\nM") - 2 * strheight("M")

Alberto Monteiro
#
On 17 Mar 2016, at 14:27 , ALBERTO VIEIRA FERREIRA MONTEIRO <albmont at centroin.com.br> wrote:

            
You still don't get it. It goes like this:
[1] 0.06733925
[1] 0.06733925
[1] 0.06733925

which appears to be identical to par("cxy")[2].