Summarised bubble plots fail when stat function requires additional arguments

[PuzzledFace]

For example:

testProp %>%
  createQtlBubblePlot(
    x=Snapshot,
    y=proportion,
    summarise=TRUE,
    stat=weighted.mean,
    w=n_patients,
    group=COUNTRY,
    size=n_patients,
    limits = list(
      list(label = "QTL (12%)", colour = "red", type = "dashed", y = 0.12, x = 1.5, vjust = -1),
      list(label = "Sec Lim (8%)", colour = "goldenrod", type = "dotted", y = 0.08, x = 3, vjust = -0.5)
    ),
    boxLeft = -2,
    boxRight = 4
  ) +
  guides(colour=guide_legend(ncol=2)) +
  coord_cartesian(xlim=c(-4, 6))
#> Error in `dplyr::summarise()`:
#> ! Problem while computing `proportion = stat(proportion, ...)`.
#> ℹ The error occurred in group 1: COUNTRY = "AUS", Snapshot = "2022-11-25".
#> Caused by error in `weighted.mean.default()`:
#> ! object 'n_patients' not found

This appears to be because weighted.average does not data-mask its arguments.

• The simplest solution is to preprocess the input dataset first [see below], but a workaround would be helpful.
• How to spot that this problem might arise during input validation?

  group_by(COUNTRY, Snapshot) %>% 
  summarise(
    proportion=weighted.mean(proportion, w=n_patients),
    n_patients=sum(n_patients),
    .groups="drop"
  ) %>% 
  createQtlBubblePlot(
    x=Snapshot,
    y=proportion,
    group=COUNTRY,
    size=n_patients,
    limits = list(
      list(label = "QTL (12%)", colour = "red", type = "dashed", y = 0.12, x = 1.5, vjust = -1),
      list(label = "Sec Lim (8%)", colour = "goldenrod", type = "dotted", y = 0.08, x = 3, vjust = -0.5)
    ),
    boxLeft = -2,
    boxRight = 4
  ) +
  guides(colour=guide_legend(ncol=2)) +
  coord_cartesian(xlim=c(-4, 6))

works as expected.