Hi Jinsong,
This is similar to the "arctext" function in plotrix. I don't want
to
do all the trig right now, but I would suggest placing the characters
on the curve and then offsetting them a constant amount at right
angles to the slope of the curve at each letter. I would first try
having a "minspace" argument to deal with crowding at small radii and
you would probably have to start at the middle and work out to each
end. A tough problem and you have made a good start on it. Check the
fragment below for a suggestion on how to avoid calling "substr"
repeatedly.
# get a vector of the characters in str
# rather than call substr all the time
strbits<-unlist(strsplit(str,""))
for(i in 1:l) {
w <- strwidth(strbits[i])
h <- strheight(strbits[i])
Jim
On Tue, Sep 22, 2020 at 6:11 PM Jinsong Zhao <jszhao at yeah.net>
wrote:>
> Hi there,
>
> I write a simple function that could place text along a curve. Since I am
not familiar with the operation of rotating graphical elements, e.g., text,
rectangle, etc., I hope you could give suggestions or hints on how to improve
it. Thanks in advance.
>
> # Here is the code:
>
> getCurrentAspect <- function() {
> uy <- diff(grconvertY(1:2,"user","inches"))
> ux <- diff(grconvertX(1:2,"user","inches"))
> uy/ux
> }
>
> r.xy <- function(o.x, o.y, theta) {
> r.x <- o.x * cos(theta) - o.y * sin(theta)
> r.y <- o.x * sin(theta) + o.y * cos(theta)
> c(r.x, r.y)
> }
>
> text.on.curve <- function(x, y, x.s, str, ...) {
>
> l <- nchar(str)
>
> fun <- approxfun(x, y, rule = 2)
>
> for(i in 1:l) {
> w <- strwidth(substr(str, i, i))
> h <- strheight(substr(str, i, i))
>
> x.l <- x.s
> x.r <- x.s + w
> y.l <- fun(x.l)
> y.r <- fun(x.r)
> theta <- atan((y.r - y.l)/(x.r - x.l) * getCurrentAspect())
>
> lb.xy <- c(x.s, fun(x.s))
> rb.xy <- lb.xy + r.xy(w, 0, theta)
> lt.xy <- lb.xy + r.xy(0, h, theta)
> rt.xy <- lb.xy + r.xy(w, h, theta)
> c.xy <- lb.xy + r.xy(w/2, h/2, theta)
>
> while(i > 1 && lt.xy[1] < rt.xy.old[1]) {
> x.s <- x.s + 0.05 * w
> x.l <- x.s
> x.r <- x.s + w
> y.l <- fun(x.l)
> y.r <- fun(x.r)
> theta <- atan((y.r - y.l)/(x.r - x.l) * getCurrentAspect())
>
> lb.xy <- c(x.s, fun(x.s))
> rb.xy <- lb.xy + r.xy(w, 0, theta)
> lt.xy <- lb.xy + r.xy(0, h, theta)
> rt.xy <- lb.xy + r.xy(w, h, theta)
> c.xy <- lb.xy + r.xy(w/2, h/2, theta)
> }
>
> x.s <- rb.xy[1]
> rt.xy.old <- rt.xy
>
> text(c.xy[1], c.xy[2], substr(str, i, i), srt = theta * 180 / pi,
...)
> }
> }
>
> # A simple demo:
>
> x <- seq(-5, 5, length.out = 100)
> y <- x^2
> plot(x,y, type = "l")
> text.on.curve(x, y, -2 ,"a demo of text on curve", col =
"red")
>
> Best,
> Jinsong
>
> [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.