You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is your feature request related to a problem? Please describe.
With the approval of sensors as groups would it make sense to move the data within its sensors(groups)? This new structure would be a platform (glider) equipped with multiple sensors, each one with its attributes and measurements. @emmerbodc , you proposed something like that on the #103 discussions. Could you be interested in leading such a PR? I would be happy to help, including providing the changes on the CDL and netCDF.
Is this related to a specific platform models
Applied to all types of gliders
Describe the solution you'd like
Move the measurements from the root level of the file to within the groups of the respective sensors. Some benefits are on reducing the change of sparse data when operating multiple sensors, and allowing equivalent measurements from multiple sensors, such as temperature from CTD + temperature from DOXY.
Describe alternatives you've considered
Keep it a flat structure as it is now.
Additional context
None at the moment.
The text was updated successfully, but these errors were encountered:
vturpin
added
question
Further information is requested
1.0
For stuff that must be resolved before we are able to release 1.0
labels
Sep 21, 2022
vturpin
added
0.1
For the stuff that we must resolve first.
and removed
1.0
For stuff that must be resolved before we are able to release 1.0
labels
Sep 30, 2022
castelao
added
1.X
For stuff to postpone and resolve only after the 1.0 release
and removed
0.1
For the stuff that we must resolve first.
labels
Apr 24, 2023
moderator: @castelao
Is your feature request related to a problem? Please describe.
With the approval of sensors as groups would it make sense to move the data within its sensors(groups)? This new structure would be a platform (glider) equipped with multiple sensors, each one with its attributes and measurements.
@emmerbodc , you proposed something like that on the #103 discussions. Could you be interested in leading such a PR? I would be happy to help, including providing the changes on the CDL and netCDF.
Is this related to a specific platform models
Applied to all types of gliders
Describe the solution you'd like
Move the measurements from the root level of the file to within the groups of the respective sensors. Some benefits are on reducing the change of sparse data when operating multiple sensors, and allowing equivalent measurements from multiple sensors, such as temperature from CTD + temperature from DOXY.
Describe alternatives you've considered
Keep it a flat structure as it is now.
Additional context
None at the moment.
The text was updated successfully, but these errors were encountered: