OceanGliders.org/GliderTools Meeting Summary

[jbusecke]

Thanks @soerenthomsen for bringing everyone together!

This is my attempt to summarise the content of the meeting on 2021/3/17 about future development on GliderTools.

Wishlist

We decided to split tasks/wishes up in three general categories:

1) Preprocessing: Stuff that has to do with different file formats and potentially platform specific steps 2) Core code and methods: Adding new features/Improving existing features 3) Documentation and Maintenance: Everything to do with the package internals and the documentation of the package

GT = GliderTools OG = OceanGliders.org

Preprocessing

• OG might not be the format that all scientific users want to use, should GT allow translation internally? Example Argopy has some conversion modules. [@dhruvbalwada ]
• Load OceanGliders Format / Produce OceanGliders format from Slocum/SeaGliders/SeaExplorer [@vturpin @dhruvbalwada]

Core code

• Diagnostic / Visualisation tools to assess data quality [@bastienqueste @marcelduplessis]
• Standardised “glider” algorithms (ie. regressions between up vs down casts) [@bastienqueste]
• better gridding modules (Dhruv) agree (IG) double agree [@marcelduplessis, SN, IG, @dhruvbalwada]
• sorting of density before mapping (people do in Matlab right now)
• How much of visualization do we want to be part of GT? I often find myself just over riding GT’s vis tools or not using them [@dhruvbalwada ]
• What are the things that are being done in these legacy MATLAB scripts that are not part of GT right now? [@dhruvbalwada]
• Standard/basic flight models (challenging…)? [@lukegre @bastienqueste,SN]
• automate splitting repeat glider transects in the horizontal space [@isgiddy]
• Co-location module cross glider/platform calibration: For commonly used auxiliary data perhaps (e,g, with ERA / surface platforms) [@isgiddy , SN, @marcelduplessis]
• Oxygen module needs to be developed further [IG, @soerenthomsen ] #74
• Gridding to density space (or maybe this falls under the gridding module) [@marcelduplessis]
• fluorescence module linked to latest developments (maybe Tommy’s work?) [@isgiddy ,SN]
• should the QC steps that are there be part of glider tools or do they go to another module? Maybe we need to strategize
• for later when OG is out about what will be done by GT and what will be done in RT (e.g despiking is done where?) [@dhruvbalwada]

  Documentation and maintenance

• Simple example notebooks on e.g. binder [@callumrollo]
• Links to best practices guide and example notebooks [@bastienqueste]
• A list of examples of common tasks we do with glider data, which can then be added to the examples that GT already has. [@dhruvbalwada ]
• Full ‘pipeline’ example. For a new user, describe where I can discover/access the data, and what steps are needed to get to a common output (section plot).[@jbusecke]
• Since often during insitu processing we make some processing choices, which might be motivated by papers or sometimes just a choice made in the moment, we should have that in the code - make some argument for why choice was made and when to change it - also helps users know why a choice is being made [@dhruvbalwada]
• [ ] Make available via conda-forge? #45 [@callumrollo @jbusecke]

Folks without github handles: Sarah Nicholson, Isabelle Giddy, Callum Rollo. If anyone knows their handles, please tag them, and Ill edit the post.

I suggest we use this issue as a central discussion place to see which of these aspects should be prioritized for the workshop.

[jbusecke]

If you think anything here is worth doing/possible to do in the given time frame, please open a separate issue. In that issue please link back to this one ( you can do that by simply typing #46).

Within the issue please describe 1) How the user would call this functionality. E.g. discuss the API 2) Outline what is needed for this feature, and if there are any roadblocks to implementing it.

[jbusecke]

To discuss if things should be combined/edited above, please use markdown quotes on the items from the original post.

For example, write

> Make available via conda-forge? [CR @jbusecke]

I ❤️ that idea and would love to implement it.

gives:

Make available via conda-forge? [CR @jbusecke]

I ❤️ that idea and would love to implement it.

[dhruvbalwada]

Thanks @soerenthomsen for organizing and @jbusecke for moving stuff over.

Maybe we move this discussion item to the discussions tab?

[callumrollo]

Thanks for organising this @jbusecke

Simple example notebooks on e.g. binder [CR]

I am happy to take on this item. Quick, easy notebooks demonstrating core functionality are the first thing I look for in a GitHub project (I'm too lazy to read the documentation). The current notebook is a bit of a monster!

[soerenthomsen]

Thanks @jbusecke! Could you edit the document above at add @isgiddy (IG) and @callumrollo (CR). Is Sarah Nicholson already on GitHub?

[jbusecke]

Maybe we move this discussion item to the discussions tab?

@dhruvbalwada I haven't used that feature at all. What is the difference/advantage of doing that?

[dhruvbalwada]

Maybe we move this discussion item to the discussions tab?

@dhruvbalwada I haven't used that feature at all. What is the difference/advantage of doing that?

Basically github has incorporated discourse like forum into itself. So non-issue discussions can live there.

[jbusecke]

Ah cool. I'd suggest we keep this one here, but I'll def start future discussions over there?

[jbusecke]

Done