Contribution of training materials

[lhmarsden]

There was some discussion after my presentation of the CF 2024 workshop that I could contribute some of the training materials I have developed to CF. Yes, I would be very happy for you to make any of these materials available via the CF website training page.

I host a YouTube channel and have written a few jupyter books with code and written explanations that accompany some of the videos. Here is a list of some of the materials that you might find relevant.

Working with CF-NetCDF in R:
Written document: https://lhmarsden.github.io/NetCDF_in_R_from_beginner_to_pro
YouTube playlist: https://www.youtube.com/playlist?list=PLpykv05eUR0oqBYFNqtiRgJGYR2tpCYtw

Working with CF-NetCDF in Python:
Written document: https://lhmarsden.github.io/NetCDF_in_Python_from_beginner_to_pro
YouTube playlist: https://www.youtube.com/playlist?list=PLpykv05eUR0rAuaLnybkZORnCwyND4CV4

Introduction to CF-NetCDF:
A short video describing the classic data format, including CF and ACDD
YouTube video: https://www.youtube.com/watch?v=FGHJhAFf1W0

If you were to endorse these materials, I would recommend that this is done with a disclaimer that these are not developed as part of CF. I can not confidently state that everything I have written or said is in-line with the best practices CF would recommend.

I also welcome any suggestions and contributions. The videos can't be edited, but the written documents are hosted on GitHub as Jupyter Books. You are welcome to submit and issue or even a pull request if you are feeling particularly generous with your time.

I don't know if this is possible with a website made using markdown etc, but there are a lot of data showing that people are much more likely to watch an embedded video than clicking on a link to a video. People also retain a lot more information from a video than from text.

I hope this is helpful, and I am happy to contribute as much or little as you like. I am also open to suggestions on making more materials that you think might be useful for the CF community.

Also, @erget has a very nice video on visualising data in NetCDF files... which has a lot more views that my videos!
https://www.youtube.com/watch?v=XqoetylQAIY

[lhmarsden]

I have moved my tutorial series (the written bits) to here:
https://www.nordatanet.no/NetCDF-in-Python-from-beginner-to-pro
https://www.nordatanet.no/NetCDF-in-R-from-beginner-to-pro

[sadielbartholomew]

Hi Luke,

I haven't had a chance to check out your training videos in any depth yet, though will certainly do in the near future, but from a glance over your materials and brief watch of some content they seem very impressive.

Speaking on behalf of myself, @japamment and others who formed a CF Training group a few years ago, with the aim to create dedicated training for the CF Conventions, we'd love it if we could include your content under our 'CF Training' page.

I think the best means to include your resources are:

• add a direct link in a bullet point under the 'CF training materials' heading to the pages you think are most relevant to link to to showcase your content, for example to the tutorial series and notebooks; and
• since we also have a dedicated YouTube channel as part of that, linked as you may have seen - is there a way we can include your videos there, whilst still allowing them to be directly attributed to your 'Luke Data Manager' user source account, in a way that you'd be happy with?

[sadielbartholomew]

I also welcome any suggestions and contributions.

In this case, I also come to you with a request. Your 'NetCDF-in-Python-from-beginner-to-pro' course looks very useful and thorough and I note the material there centres around the use of xarray to read, interpret and write out data. But as you may be aware, there are other approaches to using Python to work with netCDF, for example myself and a small team at my group, NCAS-CMS, develop a set of Python libraries called cf-python and cf-plot (amongst some others related libraries, but these two are most directly relevant to the content in your courses), which are designed for data analysis with the CF Conventions and CF Data Model at the center.

If you have the bandwidth, we'd love it if you could try out our tools, notably cf-python as a starting point, and if you find them useful ideally you could include some content where you showcase their usage as an alternative to the pure xarray way of working with netCDF, if nothing else to demonstrate that there are different Python tools for the job. We have designed these tools with CF metadata at the heart, and think therefore that they offer improved CF awareness to xarray. But, to avoid a shameless plug, I can let you make your own mind up about their usefulness 🙂 And if you find any deficiencies with these libraries relative to xarray otherwise, please let us know on the Issue Trackers - we love to get feedback and requests to help us to make the libraries as useful as possible for the community.

[lhmarsden]

• add a direct link in a bullet point under the 'CF training materials' heading to the pages you think are most relevant to link to to showcase your content, for example to the tutorial series and notebooks; and

This would be good. The materials I have listed above would be my recommendations, but feel free to pick and choose.

• since we also have a dedicated YouTube channel as part of that, linked as you may have seen - is there a way we can include your videos there, whilst still allowing them to be directly attributed to your 'Luke Data Manager' user source account, in a way that you'd be happy with?

It is possible to create a playlist of videos from other channels.

To do this:

1. Find the video you want to add
2. Click the three dots, and select 'add to playlist'
3. If the playlist doesn't exist, you will be prompted to create it. Make it public.
4. Add more videos
5. You can also go to your channel and add details about the playlist, e.g. a description.

Python libraries called cf-python and cf-plot

These sound really useful! I will try them out and certainly consider creating a video in future. Probably this will be next year as I have a bit of a backlog, but I have created a task for myself to review this.

[lhmarsden]

Speaking on behalf of myself, @japamment and others who formed a CF Training group a few years ago, with the aim to create dedicated training for the CF Conventions

Is this group still active? I would be happy to be part of this if you are looking for more contributors!

[sadielbartholomew]

Is this group still active? I would be happy to be part of this if you are looking for more contributors!

We haven't done much in the past few years, after we initiated the training, but it is an ongoing effort and open call for the inclusion of training materials (assuming they are suitable). We'd love for you to join - there is no formal process (as far as I am aware, and unless anyone corrects me, since the team started as volunteers). And perhaps your joining can prompt us to meet again to review where we are at with training...

[lhmarsden]

Wonderful! Keep me posted.

[lhmarsden]

Hi @sadielbartholomew and others.

There is a new feature in YouTube that allows you to 'collaborate' on videos. The documentation on this is quite vague for now, but essentially it allows you to have multiple 'creators' for a video. One creator uploads a video and can invite a second creator to be a collaborator. If they agree, the video will appear on their channel too, and both creators will be listed next to the video.

Do you think this would be a good way to share my videos through the CF conventions YouTube channel?

[sadielbartholomew]

Do you think this would be a good way to share my videos through the CF conventions YouTube channel?

Sounds great to me! If I recall correctly, last time we all discussed this, we ideally wanted to share videos by external creators, such as yourself @lhmarsden, under the CF Conventions channel directly, but there wasn't a good way to do this previously when they were already shared under another account, to avoid duplication and ensure proper attribution/link to the originals - does that sound right? In which case, the new YouTube feature might solve the issue fully 🙂

[lhmarsden]

Great! I will email you with an invitation link. This should be private I guess.

[lhmarsden]

To do this:

1. Find the video you want to add
2. Click the three dots, and select 'add to playlist'
3. If the playlist doesn't exist, you will be prompted to create it. Make it public.
4. Add more videos
5. You can also go to your channel and add details about the playlist, e.g. a description.

This is still an option I think for videos that you want to share via the channel but have not been 'invited' to do so/collbaborate on.

Though I think the collaboration feature is preferable where possible.