On Thu, 8 Dec 2011, fe_jasa wrote:
I replicate the CEL example for DPM [...]
I had no problem with the implementation however I encounter a problem:
apparently the time-demeaned series do not have missing values but the
original ones had.
How did you get that idea? De-meaning leaves the original
number of missing values, as is easily verified:
<hansl>
open CEL.gdt
ngd = n + 0.05
ly = log(y)
linv = log(s)
lngd = log(ngd)
eval sum(missing(ly))
eval sum(missing(linv))
eval sum(missing(lngd))
# take out time means
loop i=1..8 --quiet
smpl (time == i) --restrict --replace
ly -= mean(ly)
linv -= mean(linv)
lngd -= mean(lngd)
endloop
smpl --full
eval sum(missing(ly))
eval sum(missing(linv))
eval sum(missing(lngd))
</hansl>
Allin Cottrell