SENAITE rabbithole; Python mon amour; Alles was du wissen muss; causal inferencing; better blogging; Bayesian and other shiny new things

  • When I was previously tasked with implementing DHIS 2 for surveillance I must have spend hundreds of hours learning about it and playing with it, both in work and in my own time; the last several months I have been going through a similar process with SENAITE the laboratory information management system; I have been documenting things here as I go, such as installation, modifying functionality, styling printed reports and the myriad other steps involved in implementation; this week we should be “soft launching” it at the lab we are working with to see what they think and make any further changes required; I am confident we have chosen the best open source option, but there is further work to do around microbiology/AMR reporting, SMS reporting and integration of the data with DHIS 2.
  • Related to that LIMS work, I have been familiarising myself with the SENAITE API (WIP) and also working out how to send secure links for patients to download their own lab reports (using Bitwarden Send).
  • Not unrelated is the revival of my interest in Python: as well as reading a massive Python book and delving into the innards of SENAITE I’ve been playing with various Python apps.
  • Finally finished a massive German book that’s been on the shelves for years, and also started a good MOOC on causal inference, to follow on from a good causal inference book I finished.
  • I did a bit of work on the blog recently, upgrading it, adding a Creative Commons licence (you can choose your own licence here) and automating the process for creating RSS and Atom feeds using RSS Anything; also added buttons so you can copy code sections.
  • We actually used Bayesian statistical methods in an outbreak! and it was easier in several ways than doing it the frequentist way; I’ve been supporting another research project which is using JASP, though not in a Bayesian way.
  • Continuing to monitor interesting developments in the data science tooling space, such as Vapour, a “better R”; Positron, a “better RStudio” (more here); Typst, a “better LaTeX”; and hayagriva, a “better BibTex”.