[R] Problems with line types in plots saved as PDF files

[Pascal Oettli]

Hi,

I also see the expected lines on your pdf file with Evince (3.2.1) using 
poppler/cairo (0.18.0).

Regards,
Pascal


Le 20/02/2013 09:09, Ian Renner a écrit :
> Hi,
>
> I am trying to save a plot as a PDF with different line types. In the example below, R displays the plot correctly with 2 curves in solid lines, 2 curves in dashed lines and 1 curve as a dotted line. However, when I save the image as PDF (as attached), the dashed lines become solid. This also happens if I use the pdf command directly (by removing the # symbols).
>
> Is there any way around this? Saving the image as a JPEG or PNG file works fine but the image quality is not desirable.
>
> b.hat = 6
> a.1 = -12
> a.2 = 0
> a.3 = 200
>
> b = seq(-10, 10, 0.0002)
> l = a.1*(b - b.hat)^2 + a.2*(b - b.hat) + a.3
>
> lambda = 20
>
> p = -lambda*abs(b)
>
> pen.like = l + p
>
> y.min = 3*min(p)
> y.max = max(c(l, p, pen.like))
>
> #pdf(file = "TestPlot.pdf", 6, 6)
> #{
> plot(b, l, type = "l", ylim = c(y.min, y.max), lwd = 2, xlab = expression(beta), ylab = "", col = "green", yaxt = "n", xaxt = "n")
> points(b, p, type = "l", lty = "dotted", lwd = 2, col = "red")
> points(b, pen.like, type = "l", lwd = 2, lty = "dashed", col = "green")
>
> axis(1, at = c(0))
> axis(2, at = c(0))
>
> lambda.hat = which.max(pen.like)
> lambda.glm = which(b == b.hat)
>
> points(b[lambda.glm], l[lambda.glm], pch = 16, cex = 1.5)
> points(b[lambda.hat], l[lambda.hat], pch = 17, cex = 1.5)
>
> b.hat = -3
> a.1 = -1.5
> a.2 = 0
> a.3 = 120
>
> l = a.1*(b - b.hat)^2 + a.2*(b - b.hat) + a.3
>
> pen.like = l + p
>
> points(b, l, type = "l", lwd = 2, col = "blue")
> points(b, pen.like, type = "l", lwd = 2, lty = "dashed", col = "blue")
>
> lambda.hat = which.max(pen.like)
> lambda.glm = which(b == b.hat)
>
> points(b[lambda.glm], l[lambda.glm], pch = 16, cex = 1.5)
> points(b[lambda.hat], l[lambda.hat], pch = 17, cex = 1.5)
>
> abline(h = 0)
> abline(v = 0)
>
> #}
>
> #dev.off()
>
>
> Thanks,
>
> Ian
>
>
>
> ______________________________________________
> R-help at r-project.org mailing list
> 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.
>