collapsing to unique 'x' values in geom_violin with draw_quantiles

[bersbersbers]

I receive the warning

In regularize.values(x, y, ties, missing(ties), na.rm = na.rm) :
  collapsing to unique 'x' values

with this code:

df <- data.frame(x = 1, y = c(0, 0.25, 0.5, 0.75, 5))
ggplot2::ggplot(df, ggplot2::aes(x = x, y = y)) +
    ggplot2::geom_violin(draw_quantiles = c(0.5))

The warning does not appear with

df <- data.frame(x = 1, y = c(0, 0.25, 0.5, 0.75, 4))

so it is not a problem of the structure (including any potential duplicates) of the data.

R version 4.0.5 (2021-03-31)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 19043)

Matrix products: default

locale:
[1] LC_COLLATE=English_United States.1252
[2] LC_CTYPE=English_United States.1252
[3] LC_MONETARY=English_United States.1252
[4] LC_NUMERIC=C
[5] LC_TIME=English_United States.1252

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base

other attached packages:
[1] glue_1.4.2

loaded via a namespace (and not attached):
 [1] fansi_0.4.2      digest_0.6.27    assertthat_0.2.1 dplyr_1.0.5
 [5] utf8_1.1.4       crayon_1.4.1     grid_4.0.5       R6_2.5.0
 [9] DBI_1.1.1        lifecycle_1.0.0  gtable_0.3.0     magrittr_2.0.1
[13] scales_1.1.1     ggplot2_3.3.3    pillar_1.5.1     rlang_0.4.10
[17] farver_2.1.0     generics_0.1.0   vctrs_0.3.6      ellipsis_0.3.1
[21] labeling_0.4.2   purrr_0.3.4      munsell_0.5.0    compiler_4.0.5
[25] pkgconfig_2.0.3  colorspace_2.0-0 tidyselect_1.1.0 tibble_3.1.0

[bersbersbers]

See also https://stackoverflow.com/q/67298243/

[bersbersbers]

Here's a symmetric example.

df <- data.frame(x = 1, y = c(0, 0, 5, 10, 10))
ggplot2::ggplot(df, ggplot2::aes(x = x, y = y)) +
    ggplot2::geom_violin(draw_quantiles = c(0.5), bw = 0.1)

Also, see https://stackoverflow.com/questions/67298243/what-does-collapsing-to-unique-x-values-mean-in-this-example?noredirect=1#comment118969776_67298243

[bersbersbers]

adjust seems to be able solve this:

df <- data.frame(x = 1, y = c(0, 0.25, 0.5, 0.75, 5))
ggplot2::ggplot(df, ggplot2::aes(x = x, y = y)) +
    ggplot2::geom_violin(draw_quantiles = c(0.5), adjust=1.1)

Should the warning maybe be turned into a more useful warning hinting the user at adjust?

[bersbersbers]

I think quantitively, this only makes a difference, if ever, when the quantile is in a region of zero density, in which case its exact placement is not visible anyway. So it would be safe to use

ecdf <- stats::approxfun(dens, data$y, ties = "ordered")

to get rid of this warning, as in

create_quantile_segment_frame <- function(data, draw_quantiles) {
  dens <- cumsum(data$density) / sum(data$density)
  ecdf <- stats::approxfun(dens, data$y, ties = "ordered")
  ys <- ecdf(draw_quantiles) # these are all the y-values for quantiles

  # Get the violin bounds for the requested quantiles.
  violin.xminvs <- (stats::approxfun(data$y, data$xminv))(ys)
  violin.xmaxvs <- (stats::approxfun(data$y, data$xmaxv))(ys)

  # We have two rows per segment drawn. Each segment gets its own group.
  ggplot2:::new_data_frame(list(
    x = ggplot2:::interleave(violin.xminvs, violin.xmaxvs),
    y = rep(ys, each = 2),
    group = rep(ys, each = 2)
  ))
}
assignInNamespace("create_quantile_segment_frame", create_quantile_segment_frame, "ggplot2")


df <- data.frame(x = 1, y = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10))
ggplot2::ggplot(df, ggplot2::aes(x = x, y = y)) +
    ggplot2::geom_violin(draw_quantiles = c(0.5), bw = 0.1)

[thomasp85]

Would you be interested in providing a PR that fixes this?

[bersbersbers]

@thomasp85 sure! I still have to think about the best way to test this, but I think I know how to approach it.

[bersbersbers]

@thomasp85 #4654 is ready for review.