How to print the confidence interval and data curve via gplot to a high resolution image type in R

Question

I am trying to create high resolution images of my data using R. The loess curve and a semi-transparent confidence interval are printed with gplot. This works. When I put a wrapper around it to print an eps image, I get the following error: "Warning message: In grid.Call.graphics(L_polygon, x$x, x$y, index) : semi-transparency is not supported on this device"

The resulting *.eps image does not have the confidence interval.

I don't necessarily have to stay with gplot. I could also switch to *.tif or another high-resolution image type widely accepted by journal publications. Or I could switch from semi-transparent to a solid CI with the loess curve drawn on top. I've written a number of lines of code toward that last idea, which may eventually work, but am hoping someone knows of a more straight forward solution.

Thanks, Shawna

    ##--- HiRes 
    setwd("C:/temp")
    require(graphics)
    setEPS()
    postscript("tmp.eps", height=8, width=8)

    ##--- Plot and print the curves with their CI's
    p <- qplot(P, D, data=DF, geom=c("smooth"),
            method="loess", formula=y~x,
            group=Solution:TestDate, color=Solution, shape=TestDate,
            xlab="Pressure (mmHg)", ylab="Diameter (mm)"
    )
    p <- p + theme(panel.grid.major = element_blank(), 
            #panel.grid.minor = element_blank(), 
            panel.background = element_blank(), 
            axis.line = element_line(colour = "black"))
    p

    dev.off()


    ##--- Dataframe
    TestDate Solution    P    D
    23-Jul-13    krebs  9.6 1.38
    23-Jul-13    krebs 15.0 1.45
    23-Jul-13    krebs 20.2 1.51
    23-Jul-13    krebs 24.9 1.60
    23-Jul-13    krebs 29.9 1.68
    23-Jul-13    krebs 35.3 1.75
    23-Jul-13    krebs 40.3 1.80
    23-Jul-13    krebs 45.1 1.85
    23-Jul-13    krebs 50.2 1.89
    23-Jul-13    krebs 55.3 1.92
    23-Jul-13    krebs 50.2 1.89
    23-Jul-13    krebs 45.0 1.86
    23-Jul-13    krebs 40.1 1.83
    23-Jul-13    krebs 35.0 1.80
    23-Jul-13    krebs 30.2 1.74
    23-Jul-13    krebs 25.0 1.68
    23-Jul-13    krebs 20.1 1.60
    23-Jul-13    krebs 14.9 1.50
    23-Jul-13    krebs  9.7 1.43
    23-Jul-13    krebs  4.8 1.34
    23-Jul-13      hik 10.2 1.31
    23-Jul-13      hik 15.2 1.35
    23-Jul-13      hik 20.2 1.41
    23-Jul-13      hik 25.2 1.47
    23-Jul-13      hik 30.1 1.52
    23-Jul-13      hik 35.3 1.59
    23-Jul-13      hik 40.2 1.64
    23-Jul-13      hik 45.2 1.75
    23-Jul-13      hik 50.1 1.81
    23-Jul-13      hik 55.0 1.83
    23-Jul-13      hik 50.1 1.80
    23-Jul-13      hik 45.2 1.77
    23-Jul-13      hik 40.2 1.70
    23-Jul-13      hik 35.1 1.62
    23-Jul-13      hik 30.2 1.56
    23-Jul-13      hik 25.0 1.48
    23-Jul-13      hik 20.2 1.42
    23-Jul-13      hik 15.2 1.36
    23-Jul-13      hik 10.2 1.32
    9-Aug-13    krebs  7.5 1.48
    9-Aug-13    krebs 15.0 1.62
    9-Aug-13    krebs 20.1 1.78
    9-Aug-13    krebs 25.1 1.99
    9-Aug-13    krebs 30.1 2.27
    9-Aug-13    krebs 35.1 2.49
    9-Aug-13    krebs 40.0 2.62
    9-Aug-13    krebs 45.0 2.70
    9-Aug-13    krebs 50.0 2.74
    9-Aug-13    krebs 55.0 2.77
    9-Aug-13    krebs 50.0 2.75
    9-Aug-13    krebs 45.1 2.73
    9-Aug-13    krebs 40.0 2.71
    9-Aug-13    krebs 34.9 2.60
    9-Aug-13    krebs 30.0 2.44
    9-Aug-13    krebs 25.1 2.17
    9-Aug-13    krebs 20.1 1.90
    9-Aug-13    krebs 15.1 1.69
    9-Aug-13    krebs  9.8 1.56
    9-Aug-13      hik  9.9 1.41
    9-Aug-13      hik 14.9 1.49
    9-Aug-13      hik 20.0 1.56
    9-Aug-13      hik 25.0 1.69
    9-Aug-13      hik 30.1 1.84
    9-Aug-13      hik 35.3 2.04
    9-Aug-13      hik 40.1 2.29
    9-Aug-13      hik 45.1 2.53
    9-Aug-13      hik 50.0 2.66
    9-Aug-13      hik 55.1 2.73
    9-Aug-13      hik 50.1 2.69
    9-Aug-13      hik 44.9 2.58
    9-Aug-13      hik 40.0 2.37
    9-Aug-13      hik 35.2 2.08
    9-Aug-13      hik 30.1 1.81
    9-Aug-13      hik 24.9 1.61
    9-Aug-13      hik 19.8 1.48
    9-Aug-13      hik 14.6 1.39
    9-Aug-13      hik 10.0 1.32

Answer 1

You can use svg format (scalar vector graphics), which being vectorial always have the highest resolution at each level of zoom and support alpha transparency. Or also a pdf format.

Functions in R are svg(), pdf() which are in the package grDevice you already have in base GNU R.