Still can't make it work [🇷 for BE/BA]

posted by PharmCat  – Russia, 2020-08-07 17:42 (1777 d 08:51 ago) – Posting: # 21833
Views: 27,912

Hi ElMaestro!

❝ I tried to lump 6 subjects together at a time, and now I am getting an estimate of beta but it appears wrong, in the sense it is not corresponding to what I get for the full model.


If you calculate β with initial V - it will be wrong. It will be estimate for current V, then you calk REML end opimize θ... And then again... recalk β, ets... You can get initial β as β = inv(qr(X).R)*qr(X).Q'*y

But if you have wrong β at final iteration - it is bad case.

You can find same equation in LINDSTROM&BATES - 1.2 (where X not a design matix, but vector of Xi, IMHO)

If you provide me data from 6 subjects I can make calculations in Julia.

Magical linear algebra in julia:
using LinearAlgebra

X1 = [1 1 0 0 1; 1 0 0 0 1]
X2 = [1 1 0 0 1; 1 1 1 0 0]
Xv = Vector{Any}(undef, 2)
Xv[1] = X1
Xv[2] = X2
Xv'*Xv == X1'*X1 + X2'*X2 == vcat(X1, X2)'*vcat(X1, X2)
#true

Complete thread:

UA Flag
Activity
 Admin contact
23,424 posts in 4,927 threads, 1,677 registered users;
31 visitors (0 registered, 31 guests [including 9 identified bots]).
Forum time: 02:33 CEST (Europe/Vienna)

Complex, statistically improbable things are by their nature
more difficult to explain than
simple, statistically probable things.    Richard Dawkins

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