need to set default values for TERPSICHORE input

[pomphrey]

When running STELLOPT with a TERPSICHORE calculation of kink stability, the TERPSICHORE calculation 'hangs' if certain parameters are not set.
For example, if instead of having
TARGET_KINK(001) = 0.000000000000E+000 SIGMA_KINK(001) = 1.e-1 MLMNS_KINK(001) = 76 NJ_KINK(001) = 87 NK_KINK(001) = 47 LSSL_KINK(001) = 1537 LSSD_KINK(001) = 815
as an input line, I have
TARGET_KINK(001) = 0.000000000000E+000 SIGMA_KINK(001) = 1.e-1
(ie omitting values for MLMNS_KINK, NJ_KINK, LSSL_KINK and LSSD_KINK), the calculation hangs.
This suggests that one or all of these parameters are not set by default. I suggest they should have preset values or a warning message be sent to the user to define these quantities in the input file.

[lazersos]

The variables MLMNS_KINK, NJ_KINK, LSSL_KINK, and LSSD_KINK do have default values set in stellopt_input_mod. The problem is inside TERPSICHORE as NJ_KINK and NK_KINK depend on the number of modes set in the input file for TERPSICHORE. This could be fixed:

Project: Write subroutine which reads the TERPSICHORE input file and pre-calculates these values based on the file.

The prototype for this subroutine would look something like

TERPSICHORE_CALC_ARRAY_SIZE(FILENAME,MLMNB,NJ,NK,LSSL,LSSD,IVAC)
CHARACTER(*), INTENT(IN) :: FILENAME
INTEGER,INTENT(OUT) :: MLMNB,NJ,NK,LSSL,LSSD,IVAC
! Note MLMNV and NS are already set dynamically based on VMEC

The TERPSICHORE page can help with definitions of these variables. There's also a good PDF write-up by W. Cooper.

This would be a descent first task for a summer student, we could then remove these variables from the OPTIMUM namelist completely. It would also make TERPSICHORE more portable.