[ecco-support] "divided.ctrl" error READ operation tried to read past the end-of-file.

[dong.jian at studium.uni-hamburg.de]

Well, I just noticed a typo "delR", which was supposed to be "delZ"

Best,
Dong


Quoting dong.jian at studium.uni-hamburg.de:

> Dear ECCO community,
>
> I was running adjoint model(divided adjoint) with ecco's generic  
> cost function and got the error below, I suspect something is  
> missing in the runtime, the adjoint executable seems did not get the  
> expected input, then I did not get sensitivity.
>
>
> lib-4001 : UNRECOVERABLE library error
>   A READ operation tried to read past the end-of-file.
>
> Encountered during a list-directed READ from unit 76
> Fortran unit 76 is connected to a sequential formatted text file:
>   "divided.ctrl"
>
> I have forward outputs, snapshot and costfinal
> divided.ctrl id written as 2,  1, so it did not finish the adjoint run.
>
> my second and third divided_adjoint  output STDOUT.0000.diva_2 ended  
> with // Gridding paramters ( PARM04 in namelist )
>
>
>
> //
> (PID.TID 0000.0001) // Gridding paramters ( PARM04 in namelist )
> (PID.TID 0000.0001) //
> (PID.TID 0000.0001) usingCartesianGrid = /* Cartesian coordinates  
> flag ( True/False ) */
> (PID.TID 0000.0001)                   F
> ... ...
> (PID.TID 0000.0001) rcoord = /* P-point R coordinate (  units of r ) */
> (PID.TID 0000.0001)                -2.000000000000000E+00,      /* K =  1 */
> (PID.TID 0000.0001)                -6.500000000000000E+00,      /* K =  2 */
> (PID.TID 0000.0001)                -1.150000000000000E+01,      /* K =  3 */
> (PID.TID 0000.0001)                -1.650000000000000E+01,      /* K =  4 */
> (PID.TID 0000.0001)                -2.200000000000000E+01,      /* K =  5 */
> (PID.TID 0000.0001)                -2.800000000000000E+01,      /* K =  6 */
> (PID.TID 0000.0001)                -3.450000000000000E+01,      /* K =  7 */
> (PID.TID 0000.0001)                -4.200000000000000E+01,      /* K =  8 */
>
> Might there be something wrong with the vertical grid? (One strange  
> thing is that if I only have a few time steps, the adjoint model  
> worked well)
>
> Many thanks,
> Dong
>
>
> _______________________________________________
> ecco-support mailing list
> ecco-support at mit.edu
> https://mailman.mit.edu/mailman/listinfo/ecco-support