Bioequivalence and Bioavailability Forum • bootf2BCA v1.1

bootf2BCA v1.1 [Software]

posted by Helmut – Vienna, Austria, 2019-10-05 02:59 (1658 d 13:11 ago) – Posting: # 20665
Views: 11,788

Dear all,

we discussed the code at BioBridges…

The problem with reproducibility can easily be solved.
In the script F2_boot.R after line 270 #main algorithm add a new line

set.seed(X)

where X is an integer (specified in the protocol and report). Use

c(-1L, +1L) * .Machine$integer.max

to check the possible range. On my machine it gives

[1] -2147483647 2147483647

However, I suggest a simple one (42?).

❝ Be warned: The source code is hard to read, nearly no comments at all.

The current version 1.2 is a little bit better. Still heavy stuff.

The cut-off preferred by the EMA is implemented:

[image]

Don’t get confused by Q>=85%. That’s typo. Actually it is the correct >85%.
Still missing: One value >85% for the test (FDA) or one value >85% for the comparator (WHO).

❝ Be further warned: If you rise the number of bootstrap samples (default is 1000, too low IMHO) the result file will contain a huge number of entries, each bootstrap sample is recorded! The same mess as in the 'Object Pascal' implementation. But could of course also changed with a little knowledge of R.

Hint: In F2_boot.R comment out the loop in lines 317–329. Then the report-file shrinks from a couple of megabytes to some kilobytes. I would keep the rest for documentation. Useful to read the bootstrapped ƒ₂ values from the file and generate your own histogram (density instead of counts). I don’t like the one of

plotly. Way too many bins for my taste (here an example with 5,000 bootstraps). $\small{\textrm{ceiling}(\sqrt n)}$ gives 71 bins and smells of Excel.

[image]

I prefer the Freedman–Diaconis rule* and show additionally the median and 90% CI (green line if the lower CL ≥50).

[image]

Even if you specify hist(..., breaks="Freedman–Diaconis") this is not what you might expect from the Freedman–Diaconis rule $width=2\frac{\textrm{IQR}(x)}{\sqrt[3]{n}}$ and $bins=[\textrm{max}(x)-\textrm{min}(x)]/width$.
The man page of hist states:
[…] the number is a suggestion only; as the breakpoints will be set to pretty values.
In my example I got 34 bins instead of the desired 47 (IQR=2.9026, n=5,000, min=55.76, max=71.73). Workaround (x are the bootstrapped ƒ₂-values):
width <- 2*IQR(x)/length(x)^(1/3) bins <- as.integer(diff(range(x)) / width) hist(x, breaks = seq(min(x), max(x), length.out = bins), freq = FALSE)

—
Dif-tor heh smusma 🖖🏼 Довге життя Україна!
Helmut Schütz

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

Complete thread:

Validation of PhEq_bootstrap Helmut 2018-07-12 17:47 [Software]
- Validation of PhEq_bootstrap ElMaestro 2018-07-12 22:21
  - Validation of PhEq_bootstrap Helmut 2018-07-13 01:29
    - Validation of PhEq_bootstrap ElMaestro 2018-07-13 10:14
      - Validation of PhEq_bootstrap Helmut 2018-07-13 12:54
        
        Validation of PhEq_bootstrap ElMaestro 2018-07-13 14:28
- Validation of PhEq_bootstrap Shuanghe 2018-07-17 15:54
- PhEq_bootstrap in R d_labes 2018-08-15 09:45
  - bootf2BCA v1.1Helmut 2019-10-05 00:59
    - bootf2BCA v1.2: seed Helmut 2020-09-25 19:37
      - bootf2BCA v1.2: seed PTM139 2021-04-16 07:53
        
        Which packages, working directory? Helmut 2021-04-16 10:50
        
        Which packages, working directory? PTM139 2021-04-27 13:02
      - bootf2BCA v1.2: seed Rayhope 2022-03-28 14:35
        
        bootf2BCA v1.3 Helmut 2022-03-28 14:48
        
        bootf2BCA v1.3 Rayhope 2022-03-30 06:34
    - bootf2BCA v1.1 Flam 2023-06-21 03:41
      - bootf2BCA v1.3.1 (current since 2022-06-25) Helmut 2023-06-23 13:47
        
        bootf2BCA v1.3.1 (current since 2022-06-25) Flam 2023-06-26 23:16
        
        Install required packages first Helmut 2023-06-27 11:47
- Validation of PhEq_bootstrap Rayhope 2021-07-29 08:49
  - Validation of PhEq_bootstrap Helmut 2021-08-02 18:12