Draft: Resolve "Add dataset_descrtion IDS"

assigned to @lsimons

added 1 commit

8ff7659f - Added get_mds_node_info.m and tcv_get_ids_dataset_description.m

requested review from @mmarin

I have attempted to fill this IDS but I need some help to decide how to fill a few fields:

We have the choice to set whether the ids is filled with experimental or simulation data. I'm not sure how best to determine this. @osauter perhaps you can help? I put this as a place holder for now:

% FIXME: Not sure what the condition should be here... To be discussed
if strcmp(gdat_params.data_request,'ids')
    ids_dataset_description.type.name='experimental';
    ids_dataset_description.type.index=1;
    ids_dataset_description.type.description='Data coming from an experiment';
else
    ids_dataset_description.type.name='simulation';
    ids_dataset_description.type.index=2;
    ids_dataset_description.type.description='Data coming from a simulation';
end

For filling information about the pulse time, I copied the example of @yildiz but I'm not sure how to get the pulse_time_end information.

% Attempt to fill information about pulse time
% FIXME: How to get pulse_time_end? Simply = pulse_time_begin + X seconds?
[pulse_time_begin,pulse_time_begin_epoch]= ...
    get_mds_node_info('\VSYSTEM:TIMERS.VALUES',shot);
ids_dataset_description.pulse_time_begin=pulse_time_begin;
ids_dataset_description.pulse_time_begin_epoch.seconds=pulse_time_begin_epoch;
ids_dataset_description.pulse_time_begin_epoch.nanoseconds=pulse_time_begin_epoch.*1e9;
% ids_dataset_description.pulse_time_end_epoch.seconds=
% ids_dataset_description.pulse_time_end_epoch.nanoseconds=
% ids_dataset_description.pulse_processing_time_begin=node_time

There is some information related to simulation quantities. I'm not sure how we are supposed to fill this information:

% FIXME: How do we hope to get this information?
% ids_dataset_description.simulation.comment_before=?

Are these all required properties? I don't think we have to strictly fill every property, no? Even starting with ones mentioned in #75 would be a good start?

But if you must fill them, some comments:

ids_dataset_description.pulse_time_begin=pulse_time_begin;

pulse start is not exactly the time given by TIME_INSERTED property of this node, but I don't think we have anything closer.

ids_dataset_description.pulse_time_begin_epoch.nanoseconds=pulse_time_begin_epoch.*1e9;

This is ns component of the epoch seconds (it's between 0 and 1E9)

https://crpplocal.epfl.ch/~sauter/IMAS/dataset_description.html

There is some information related to simulation quantities. I'm not sure how we are supposed to fill this information:

As we run gdat on real tcv_shots, I assume it's never a simulation, but always "pulse" (experimental), right?

These aren't required proprerties we can just open a separate issue and fill them later. Here's a new issue: #79 to keep track of that.

pulse start is not exactly the time given by TIME_INSERTED property of this node, but I don't think we have anything closer.

Okay well in that case I guess TIME_INSERTED will be the best choice.

This is ns component of the epoch seconds (it's between 0 and 1E9)

Ah I see, I misunderstood that, I will correct it now

Fixed ids_dataset_description.pulse_time_begin_epoch.nanoseconds. Just waiting on input from Olivier before resolving this thread

mentioned in issue #79

added 1 commit

d106e4b8 - Changed comments. Added calculation for epoch time in ns

Compare with previous version

Draft: Resolve "Add dataset_descrtion IDS"

Merge request reports

Activity