Missing values in @logl series

[drivenbark]

Experts,

I’m new to EViews programming and my question may seem trivial to most of you, I apologize for that. I’ve used the sample program hetero1.prg as a template to estimate a similar model. The key difference are as follows:

(1) I have used the “fetch” command and to get levels of my dependent data series from a previously saved database.
(2) I used the dlog(“y”,1,52) command to take the normalize my data and remove both seasonality and trend, the data is weekly.
(3) My independent variable is a dummy variable.

When I run my model I receive the error message:

Missing values in @logl series at current coefficients at observation 9/11/2000 in “do_ll2807.ml(showopts,m=1000,c=1e-5).”

I think it has something to do with the “dlog” command. Can anyone direct me to a document that would explain what causes this error? I haven’t had any success searching the internet. My program is attached below.

Thanks,
David

'create workfile
create “test_work” w 1/4/1999 10/17/2009

'fetch data
'***note to self -- no spaces between fetch and (d=...)
fetch(d=fdrevs) sr2807

'take log year over year difference
series y2807=dlog(sr2807,1,52)

'set the dollar general open variable
series dg = 0
dg.fill(o=9/5/2005,l) 1

'specify the likelihood functions for store #2807, the primary interest
logl ll2807
ll2807.append @logl logl2807
ll2807.append res2807 = y2807-c(1)-c(2)*dg
ll2807.append var2807 = c(3)*dg
ll2807.append logl2807= log(@dnorm(res2807/@sqrt(var2807)))-log(var2807)/2

'set starting values
equation eq2807.ls y2807 c dg
for !i=1 to 2
c(!i)=eq2807.c(!i)
next
c(3) = eq2807.@se^2

'estimate for sample with nonmissing values for dependent variable
smpl @all if y2807<>na
ll2807.ml(showopts, m=1000, c=1e-5)

'print the output
show ll2807.output