accelerated time model with proportional hazards

#1
$coef
Estimate SE
lambda 0.004424690 0.005026449
gamma 0.888627728 0.108972639
treatment -0.646997631 0.309927354
gender -0.701293522 0.392452782
race 2.105048640 1.015077273
percentage 0.003336242 0.008991924
head -0.049555562 0.342919026
buttocks 0.610699987 0.414709100
trunk 0.117181285 0.487344728
upperleg -0.446654667 0.375326008
lowerleg -0.262843680 0.378571921


(Intercept) 6.09992 1.2876 4.737 2.16e-06
treatment 0.72809 0.3606 2.019 4.35e-02
gender 0.78919 0.4437 1.779 7.53e-02
race -2.36888 1.1702 -2.024 4.29e-02
percentage -0.00375 0.0102 -0.369 7.12e-01
head 0.05577 0.3861 0.144 8.85e-01
buttocks -0.68724 0.4694 -1.464 1.43e-01
trunk -0.13187 0.5499 -0.240 8.10e-01
upperleg 0.50263 0.4236 1.187 2.35e-01
lowerleg 0.29579 0.4257 0.695 4.87e-01
Log(scale) 0.11808 0.1226 0.963 3.36e-01
Scale= 1.13

Accelerated failure time model
Log(T)=y= μ+ α^T z+ σW

Proportional hazards model
h(x│z)=(γλt^(γ-1) )exp(β^T z)


γ=1/σ
λ=exp(-μ/σ)
β=-α/σ

Problem if i compute exp(-6.09992/Scale= 1.13) it should be 0.004424690

But it isnt :(

What do i do wrong ?
What do i do wrong ?

Is the significance the same for the computed beta's ?
 
Last edited: