Displaying 1 result from an estimated 1 matches for "pumps_ti".
2004 Nov 16
1
lme, two random effects, poisson distribution
...approximated by
Poisson dist. (# of pumps is very often = 0)
The slugs were observed during 12 time slots which are correlated in
time as AR(1). The time slots are divided into two categories:
Resting time slots (the first 10)
Excited time slots (the last 2)
I used model:
************pumps_ti = state_t + slugs_i + error_ti
slugs and error are normaly distributed
pumps_ti - # of pumps for i-th animal and t-th time slot
x_t - order of the time slot (x_1 = 1, ..., x_12 = 12)
state_t - state_t = 0 for resting time slots (t=1,...,10)
state_t = 1 for excited...