compound variables

[Katia Fernandes]

Dear Grads experts,

I am trying to calculate correlation between 2 variables,
one say is temperature and the other is RH. The dimensions are
the same for both variables: daily data (t=730, 2 non-leap years), Xdef 31,
Ydef 29, Zdef 1.

But, I want to calculate the correlations month by month, so for example
I want the correlation for September (60 days sample) without having
to fwrite monthly files.

The way I tried to approach it was like the following:


*at this point I will have the times corresponding to the first and last
days of Sep. The count will be 60, so I would have 60 variables
ttsep.1...ttsep.60, same for rhsep.

**********************************
t=t1
while(t<=t2)
    count=count+1
    say count
   'set t 't
    ttsep.count=temp
    rhsep.count=rh
    t=t+1
endwhile


Assuming the above is correct, I got stuck on what I think is a trivial
problem. I now have 60 variables for both temp and RH. Can I calculate
the "time" correlation between them using the built in Grads functions?

Thank you,

Kátia Fernandes