Visualization [GxP / QC / QA]

posted by Helmut Homepage – Vienna, Austria, 2021-09-18 15:26 (250 d 23:09 ago) – Posting: # 22588
Views: 2,356

Hi ElMaestro,

» […] it makes very good sense to plot for example the RMSE, CV, SE of diff., or even the width of the CI as function of (cumulated) number of subjects.

If you have all data, of course. A simple visualization using the res data.frame of my previous post:

[image]




windows(width = 5.5, height = 5.5)
op <- par(no.readonly = TRUE)
par(mar = c(3, 4, 2, 0), cex.axis = 0.9, cex.main = 1, font.main = 1)
split.screen(c(length(CRO), length(metric)))
scrn <- 0
for (j in seq_along(CRO)) {
  for (k in seq_along(metric)) {
    scrn <- scrn + 1
    screen(scrn)
    plot(c(0.5, length(study)+0.5), rep(100, 2), type = "n",
         ylim = range(res$lower[res$CRO == CRO[j] & res$metric == metric[k]],
                      res$upper[res$CRO == CRO[j] & res$metric == metric[k]]),
         log = "y", xlab = "", ylab = "PE, 90% CI", axes = FALSE,
         main = paste0(CRO[j], ": ", metric[k]))
    axis(2, las = 1)
    axis(1, at = seq_along(study), tick = FALSE,
         labels = paste("n =", res$n[res$CRO == CRO[j] &
                               res$metric == metric[k]]))
    abline(h = c(80, 100, 125), col = c("red", "black", "red"))
    box()
    for (l in seq_along(study)) {
      arrows(x0 = l,
             y0 = res$lower[res$CRO == CRO[j] & res$metric == metric[k] &
                            res$study == study[l]],
             y1 = res$upper[res$CRO == CRO[j] & res$metric == metric[k] &
                            res$study == study[l]], length = 0.08,
             angle = 90, lwd = 2, code = 3)
      points(l, res$PE[res$CRO == CRO[j] & res$metric == metric[k] &
                       res$study == study[l]], pch = 19, cex = 1.25)
    }
  }
}
close.screen(all = TRUE)
par(op)


Dif-tor heh smusma 🖖 [image]
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes

Complete thread:

UA Flag
Activity
 Admin contact
22,108 posts in 4,630 threads, 1,567 registered users;
online 10 (0 registered, 10 guests [including 3 identified bots]).
Forum time: Friday 14:36 CEST (Europe/Vienna)

We absolutely must leave room for doubt
or there is no progress and no learning.
There is no learning without having to pose a question.
And a question requires doubt.
People search for certainty.
But there is no certainty.    Richard Feynman

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5