Efit With Pcs Data

John Ferron, September 1997

Changes made to module getecdx.for:

1. Add namelist in4 which contains the variables that will be read from the pcsnames.dat file. When this namelist is read some of the values read from dprobe.dat and some of the values hardwired into the code with data statements are replaced.
2. After reading dprobe.dat, optionally read the namelist in pcsnames.dat. Dprobe.dat must be read also because it contains more information than pcsnames.dat. Pcsnames.dat simply replaces the pointnames but the information about probe location etc. is the same.
3. The variables use_alternate_pointnames, alternate_pointname_file and do_spline_fit should be allowed to be in the snap file. These variables should be passed into the subroutine getpts. The default for alternate_pointname_file should be pcsnames.dat. If alternate_pointname_file is in the snap file it can specify another file name that replaces the default. If use_alternate_pointnames!=0 then the alternate point name file is read.
4. To get the plasma current from the pcs data requires something special because the plasma current is not explicitly in a pointname yet.
   • f use_alternate_pointnames = 1 the plasma current is computed using the pr15v pointname which is the same as ip except that the inherent number is not in the database correctly. The problem with this method is that the bit resolution is not defined for this pointname.
   • if use_alternate_pointnames = 2 the rogowskii loop and loop voltage are used to compute ip using the standard formula. In this case nsingl(5) must be rl03 (Rather than the normal rl01) and nsingl(3) must be vloopb (rather than the normal vloop).
5. The subroutine getdat was completely removed. A new version which is the same as Kathy's with a new entry point for efit added was placed at the end of the file. Eventually this routine should be removed and efit should be linked with Kathy's shareable library.
6. All calls to getdat replaced with calls to getdat_e which is the efit entry point to Kathy's routine.
7. If the pointname is NONE then no attempt is made to get data. Instead the flag is set to indicate that an error occurred trying to get this data. This is for pointnames that are known to not exist so that time is not wasted searching for them.
8. An option was added to turn off spline fit of the data. This uses the variable do_spline_fit.
9. The getdia routine needs to be rewritten to use the standard efit data accessing routines. Since the pointnames are hardwired into the diamag routine, all diamag compensation is still done using standard pointnames.
10. The pcs does not yet have all of the density pointnames. For now the standard pointnames are used.
11. The pcs pointname for bti322 is not working yet I think. For now the standard pointname is used.
12. The pcs does not have pinj.
13. It is necessary to link with -lrevshare.