[ADMB Users] indexing with a 3-years-lag

[Ben Bolker]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 10-12-07 08:25 AM, Luis Ridao wrote:
> ADMB-users,
> 
> I'm having problems to fit a ssb-rec Ricker curve.
> The problem is with the rec array indexing.
> Recruitment strength is measured at age 3 so what I do is to create
> two extras arrays holding the ssb-rec pairs with a 3-yrs lag.
> I thought this solution would make it easier but I get the following
> error:
> 
> LRC $ ./catage6109_rec_mod_surv_ric
> ok to here
> Incompatible bounds in dvar_vector& dvar_vector::operator = (_CONST
> dvar_vector& t)
> 
> Snipet of the tpl file:
> 
> nyrs=49
> .......................
>   sdreport_vector ssb(1,nyrs)
>   sdreport_vector ssb_m3(1,nyrs-3)      // SAME AS SSB with dimension (1,46)
>   sdreport_vector rec(1,nyrs)
>   sdreport_vector rec_m3(1,nyrs-3)      // SAME AS REC with dimension (1,46)
>   sdreport_vector pred_rec(1,nyrs)
>   sdreport_vector pred_rec_m3(1,nyrs-3)    // SAME AS PRED_REC with
> dimension (1,46)
> .........................
> 
> FUNCTION ssb_rec_ricker
>   ssb=rowsum(elem_prod(elem_prod(N,weight_at_age),mat_at_age))/1000;
>   ssb_m3=ssb.sub(1,46);
>   rec=column(N,1)/1000;
>   cout<<"ok to here"<<endl;          // HERE IS THE
>   rec_m3=rec.sub(3,49);              // PROBLEM
> 
>   for (int i=1;i<=nyrs-3;i++)
>   {
>   pred_rec_m3(i)=exp(log_alfa)*ssb_m3(i)*exp(-exp(log_beta)*ssb_m3(i));
>   } 
> 
> ............................
> 
> The valid index for rec is (1,49)
> Why can't I subset it to (3,49)?
> 

  I think your problem may be that rec.sub(3,49) should be 47 elements
long, but the vector you have allocated is only 46 long ... ?
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAkz+RCEACgkQc5UpGjwzenNSXACeK0LuOXreHDgPxz5/zKmydJcR
oD8AnjlSGLDqFmO7ymMAVnpuzsa96DpZ
=B90p
-----END PGP SIGNATURE-----