qicharts: is there a way for different facets of a chart to calculate central line and limits on different parts of the data?

In the qic() function of the R package qicharts2 (v0.7.5), is there a way to map the "part" argument so that the data used for calculating central line and limits can vary by facet?

I modified the following example, taken from the package's vignette, by appending a "part" argument of length 2 with intent that the 1st element would apply to the 1st facet and the 2nd element to the 2nd facet. As expected, qic() applies both elements to both facets. I'm looking for a way around that.

#From example "Faceting readmission rates by gender"

library(qicharts2)

cabg <- cabg %>% mutate(month = as.Date(cut(date, 'month')))

cabg_by_month_gender <- cabg %>% 
  group_by(month, gender) %>% 
  summarise(readmissions = sum(readmission),
            n = n())

qic(month, readmissions, n,
    data      = cabg_by_month_gender,
    facets    = ~ gender, 
    chart     = 'p',
    y.percent = TRUE,
    title     = 'Readmissions within 30 days (P chart)',
    ylab      = '',
    xlab      = 'Month',
    part      = c(10, 20)  # Argument I added
)

Solution

I couldn't find a way through the library qicharts2, but I do have a work around.

Instead of one plot make two, each with only 1 facet's data.
Even though there is only one facet's data, still declare the facet so that you get the strip text (gray background label with the group title).
Only the first should have a title.
If the y-scale is different, as is with the example data, set the y-axis limits.

This uses the package patchwork.

library(qicharts2)
library(patchwork)

p1 <- qic(x = month, y = readmissions, n = n,
    data      = cabg_by_month_gender[cabg_by_month_gender$gender == "Female", ],
    facets    = ~ gender, 
    chart     = 'p',
    y.percent = TRUE,
    title     = 'Readmissions within 30 days (P chart)',
    ylab      = '',
    xlab      = 'Month',
    part      = 10)          # <---- only the part particular to this group


p2 <- qic(x = month, y = readmissions, n = n,
    data      = cabg_by_month_gender[cabg_by_month_gender$gender == "Male", ],
    facets    = ~ gender, 
    chart     = 'p',
    y.percent = TRUE,
    title     = NULL,        # <---- I changed!
    ylab      = '',
    xlab      = 'Month',
    part      = 20) +        # <---- only the part particular to this group
  scale_y_continuous(label = scales::percent, limits = c(0, .8)) # <--- I'm new


p1 + p2 + plot_layout(axes = "collect_y")