Should label placers be evaluated at drawing time?

[eliocamp]

The placer function computes the angle of the text, so that it follows the isoline, but since it's computed when creating the isoline grob and using the raw x and y values, it doesn't work for arbitrary aspect ratios and / or when changing aspect ratio.

From the example:

library(isoband)
library(grid)

x <- (0:(ncol(volcano) - 1))/(ncol(volcano) - 1)
y <- ((nrow(volcano) - 1):0)/(nrow(volcano) - 1)
lines <- isolines(x, y, volcano, 5*(19:38))



g <- isolines_grob(
    lines, breaks = 20*(5:10),
    label_placer = label_placer_middle(),
    gp = gpar(
        fontsize = 10,
        lwd = c(1, 2, 1, 1),
        col = c("grey50", "grey20", "grey50", "grey50")
    )
)

grid.newpage()
grid.draw(g)

I get

Notice that, for example, the 140 text doesn't perfectly follow it's isoline.

Furthermore, the angles remain constant if I change the aspect ratio of the graphics device.

In metR geom_text_contour() evaluates the angle at draw time, so the angle of the text changes as the aspect ratio changes.

library(ggplot2)
ggplot(reshape2::melt(volcano), aes(Var1, Var2)) +
    geom_contour(aes(z = value)) +
    metR::geom_text_contour(aes(z = value),
                            label.placement = label_placement_fraction())

[eliocamp]

While #28 solves the issue of changing aspect ratio, I think there are other issues with not computing placement at draw time. One direct application of selecting placement at draw time is labelling on the "flattest" segment. I do something like that with metR::label_placement_flattest, which places labels the segment with minimal curvature. I could be wrong (i haven't done the math or anything) but it seems to be that the least "curvy" segment will change depending on aspect ratio.