Displaying 1 result from an estimated 1 matches for "beta_12".
Did you mean:
beta12
2010 Jul 21
0
Piecewise regression using lme()
...=1:20+rnorm(20,0,2)
segment2=20-2*1:30+rnorm(30,0,5)
segment3=-40+0.5*1:15+rnorm(15,0,1)
group=c(rep(1,20),rep(2,30),rep(3,15))
y=c(segment1,segment2,segment3)
Data=data.frame(y,t=1:65,group=as.factor(group))
the model I'd like to fit is:
y_t=
(beta_01+beta_11*t+error_1)*(group==1)+(beta_02+beta_12*t+error_2)*(group==2)+(beta_03+beta_13*t+error_3)*(group==3)
It looks like a mixed effects model were the fixed effect are the piecewise
linear regression parts (beta_0i+beta_1i*t) and the random effects are the
variance components error_i.
The problem is that I can't find the way the write...