Fix get_z_parameters for DIII-D

[yumouwei]

Problem

• zcur and zcur_normalized failed testing with data (mismatch with SQL table).

Analysis of the problem

1. zcur signal is fetched from PTDATA: VPSZP. I took a look at that pointname in ReviewPlus and the signal looks identical to the output of disruption-py.

In the docstring of the MATLAB script it mentioned that they previously used a wrong pointname and corrected it afterward. I wonder if the stored SQL table data was fetched from that address.

disruption-py/DIII-D/get_Z_error_d3d.m

Lines 29 to 31 in 5a13906

	% Revision history
	% 2019/09/04 - RSG; We found out from Jayson Barr that we misunderstood
	% the PCS pointnames. The correct pointname is "vpszp".

Implemented changes

• Broke down nested try-except block.
• Added docstring.

Problem to be addressed

• Right now z_prog, z_error and z_error_normalized are always returned as arrays of nan due to the quoted reason below. Should we keep these parameters or remove them from the outputs?

disruption-py/DIII-D/get_Z_error_d3d.m

Lines 32 to 37 in 5a13906

	% 2019/09/23 - RSG; On DIII-D the plasma control system uses isoflux
	% control to control the plasma shape and position. It does
	% NOT use zcur control. Therefore, the PCS does not have a
	% programmed vertical position. This this routine will now
	% always return an arrays of NaN for z_prog, z_error, and
	% z_error_norm.

[gtrevisan]

should I go ahead and merge?

[yumouwei]

Go ahead but we need to remember to discuss the point with z_prog, z_error and z_error_normalized with Cristina