A small graveyard
I had, for some time, some shiny apps that I made by way of a portfolio. R Shiny was (and still is) a tool for making quick-and-dirty dashboards and other interactive apps backed by the statistically-focused R programming language. You can also make much more polished web apps, and I have done so, but I find that after a certain point the shiny framework goes from helping to hindering - your milage may vary, and I still use and appreciate the system.
It’s also a bit hefty to run and so my apps (which were all self-hosted) have been turned off for a while. As such the links below may not work. I’m hoping to do some maintinence, moving the server they’re on etc, and get them back up and running again for posterity, but that is a future thing. And then of course they will probably break or need updating at some point in the future again. I should just put them up on shinyapps.io and let them do the sysadmin work - but then, if I were that way inclined this site would be quite different.
Some of the text below is taken from info pages as they appeared on the old version of this website; others I’ve written about now that I didn’t get around to at the time.
NZ Greenhouse Gas Emissions
This was an R Shiny app that takes data from this Ministry of Economic Development dataset to display the amount of greenhouse gases (in kt CO₂ equivalent) produced by New Zealand from 1990 to 2010.
Graphs could be created by sector or by fuel. The git repository for this app can be found here.
Gender Reassignment Surgery
This shiny app collated the data obtained by other people about the New Zealand GRS waiting list prior to the 2018 funding changes, and allows the user to project what would have happened without a change into the future. As it stood at the last OIA request prior to the change the vaginoplasty waiting list was 51⅓ years long, while the phalloplasty list was 52 years long—both were swiftly rising.
This was a very simple model, just integrating current totals and rates into a projection, but it raised an important point. Now that nearly 5 years has passed I should come back to this issue.
View the app here; find the code here. I also wrote a blog about this one at the time which can now be found here.
GP Retirement
Similar to the GRS app this model looked at the 2017 RNZCGP annual survey of GPs, and projected their numbers assuming stated intentions to retire would be followed through. It allowed making certain assumptions about new trainees coming to replace them.
There and Back Again
I forget all the details but in 2020 (I think) Stats NZ held a competition for an interactive visualisation of commuter data from the 2018 Census called There and Back Again. I did not win, but I did learn a lot both about making a polished app (this one was much better than the others) and also about making maps in shiny specifically. This has been very useful in other projects, but those were for work so they’re not here.
This is by far my favourite of these projects, and the one I most want to restore to proper function. In particular I ensured that it had keyboard shortcuts but could also run on lower-end phones (like the one I had at the time) - I could have sworn this was part of the original brief but was dropped at some point.
The link for this one, when it works, is here, but you should be able to make your own version from the git repository.