2024|20: Green View, the Art of Training and data gardening
Street View Green View
In March, I contributed to Street View Green View, an American Red Cross project that classifies an area by the amount of greenery visible in Street View imagery. This week, I worked on some code to process the green index scores generated by package so they can be displayed on a webmap.
I used a h3 hex grid to summarise the data, and MapLibre for the interactive webmapping. While h3 has a Python package (`h3`), I used the `h3-geopandas` package, which makes it just that bit easier to pass data to and from geodataframes.
Webmapping relies a lot on JavaScript, and MapLibre is no different. While I know the basics, I’m not an expert, so I relied on Python to set up variables and then passed them into the HTML doc via the Jinja templating engine. This turned out to be a great shortcut — I got the same end effect but didn’t have to spend a lot of time learning new bits of JS.
IFRC Art of Training course
As part of my effort to better understand best practices in facilitating training, I took IFRC’s Art of Training online course. It has four 40 minute modules on facilitation, mentoring and assessment. The bits on mentoring were less relevant to me, but I found the sections on training design and facilitation really helpful.
Data gardening
I spent some time ‘gardening’ some of the datasets we host for others in the organisation to use: adding new data where it’s available, fixing some minor errors, and deprioritising some data-related Jira tasks that have been on the backlog for a while (after talking to colleagues and finding out they’re no longer needed).
