Bioequivalence and Bioavailability Forum • Geometric mean? re: Bear: feature requests

Geometric mean? re: Bear: feature requests [🇷 for BE/BA]

posted by yjlee168 – Kaohsiung, Taiwan, 2013-03-27 14:21 (4835 d 06:07 ago) – Posting: # 10294
Views: 13,881

Dear Helmut & all,

Yes, I am doing this part. However, the question is all data points cannot be zero (the first data point is the case?) or strictly speaking all data points must greater than 1.0 (otherwise, it will cause negative values, such as log10(0.65)). It may be quite common to have some Cp < 1.0 in BE study. In R, there is some packages that can ignore that, such as the option 'zero.rm(foo)'; however, these packages are still not available for R3.0.0. Any suggestion? Thanks in advanced.

❝ • Think about geometric means/SD; some (European) regulators prefer a plot of geometric rather arithmetic means.

❝ Hint: SD_geom=exp(sqrt(1/(n-1) × Sum((ln(x)-ln(x_geom))²)) - hope, I got all the parentheses right.

—
All the best,
-- Yung-jin Lee
bear v2.9.6:- created by Hsin-ya Lee & Yung-jin Lee
Kaohsiung, Taiwan https://www.pkpd168.com/bear
Download link (updated) -> here

Complete thread:

Bear: feature requests Helmut 2010-12-08 15:48
- Bear: feature requests yjlee168 2010-12-08 18:49
  - Bear: feature requests Helmut 2010-12-08 19:48
  - Bear: feature requests d_labes 2010-12-09 08:48
    - Bear: feature requests Helmut 2010-12-09 14:20
      - Bear: feature requests yjlee168 2010-12-09 20:24
    - Bear: feature requests yjlee168 2010-12-09 19:57
      - Bear: feature requests Aceto81 2010-12-13 10:26
        
        Bear: feature requests yjlee168 2010-12-13 13:29
- Geometric mean? re: Bear: feature requestsyjlee168 2013-03-27 13:21
  - Geometric mean in plots Helmut 2013-03-27 14:18
    - Geometric mean in plots yjlee168 2013-03-27 21:16