GHSCI familiarisation session - 2024-06-05

[carlhiggs]

We had a familiarisation session for recent updates to our software on Wednesday 5 June (Melbourne time) with members of the study executive and spatial team in attendance.

The idea of session was to provide a hands on opportunity to work with our tools as they currently are to get a shared appreciation of how these can be used to calculate and report on policy and spatial indicators, and how we might improve them, and we'll have a follow up workshop on this. The idea was to take our time and run through the demonstration materials together.

I believe we achieved what was the main aim from my perspective, to run through recent updates and changes in using the software, so people were aware of them.

However, there were some clashes of expectations in terms of level of content; some felt it too basic, but then again there were people from a range of backgrounds in the group, some who had never used the software, so if we had conducted the workshop in a different way, others may have been alienated.

There were great suggestions for how we could conduct future workshops:

• Encourage users to get set up and run through basic workflow ahead of time, so workshop deals with more advanced topics or configuration issues (thanks @dapugacheva).

  • Assist attendees with issues faced prior to workshop if they report on these
    • For example how to deal with administrative rights for Docker installation (as per Experience evaluation GHSCI v4.4.10 #375 (comment) ; Docker installation instructions guide this, but we should guide our users towards those directions in our FAQ)
  • Share with the group any issues attendees reported prior to the workshop
  • In workshop, users can then run through analysis of their own study region

• Have seperate workshop topics (thanks @gboeing )

  • modern data science primer (assumed knowledge)
  • our software
    • although, I expect it still will be the case that for use of our software we will need to provide different kinds of materials for different users and perhaps there is a place for discussing modern data science workflows in the context of the software (encouraging users to develop a configuration file like an analysis plan before conducting analysis,generate documented resources in a reproducible way and iteratively validate before disseminating to a broad range of audiences is at the heart of this software, so demonstrating the software is inherently tied with encouragement of modern data science practices)

Good suggestion to have seperate training videos on these aspects.

Those were the main points covered in the workshop, other than the run through itself in which I demonstrated

• new Wiki (https://github.com/healthysustainablecities/global-indicators/wiki) and FAQ

  • we can update and migrate our existing software documentation and FAQ there

• changes to the GHSCI app, including

  • study region boundaries on map
  • indicator summaries
  • generate interactive spatial distribution maps for indicators

• changes to Jupyter notebook

  • new and simplified code syntax
  • get help on GHSCI functions (ghsci.help())
  • load example region helper (r = ghsci.example())
  • get help on GHSCI region object functions (r.help())
  • generate new reports
  • additional explanations

I demonstrated all this in the context of running the example analysis.

I also demonstrated and encouraged people to use the Jupyter Lab code editor for their configuration files to make YAML editing easier, which for users not familiar with this may have been helpful.

The video for the session is here. Currently, anyone can view this I believe but only for a month (until late July 2024). We should download and host somewhere proper other than RMIT Microsoft Stream for posterity if people are interested in it as a resource.