Delighted to share the (preprint) output of @mhairijan.bsky.social's Oxford visit:
SARS-CoV-2 Neutralising Antibody Profiles Reveal Variant Specific Antibody Dynamics and Regional Differences in Infection Histories in Malawi. dx.doi.org/10.2139/ssrn...
🎓 One last PhD studentship on offer on health analytics and modelling, jointly between @hpruham.bsky.social and @hprurespinf.bsky.social. Real-time use of data from household studies to inform pandemic and epidemic preparedness, with Anne Cori, @seabbs.bsky.social and Chris Overton. Link ↓
Budding infectious disease modellers:
We're running our intermediate-level infectious disease modelling short courses at the London School of Hygiene and Tropical Medicine. Each course is 4 days, and we're running them over the first two weeks of September.
An example of a related project is cran.r-project.org/web/packages...
which tries to do this for forecast objects but its quite an uphill struggle.
For me it seems like some nice connectors in the first instance would be great then perhaps worth thinking about if there is enough shared structure to make a meta object for targetting common tasks at.
pretty much always miss the mark i.e. the incidence family of packages.
perhaps that isn't what you are suggesting?
Hmm I am not really seeing the upside to integrating these from having a look. It looks like they are being treated similarly but with domain related differences you might expect. I am pretty leery of trying to create shared objects that are used by modelling packages as in domain they seem to
Looking at this in more detail there is a lot of overlap between these two packages.
I find the {reviser} @ropensci.org 📦 by Mark Burri + Philipp Wegmueller interesting. It shows how economists deal with a problem that epidemiologists grapple with a lot: revisions of past time series ropensci.org/blog/2026/04... #rstats @seabbs.bsky.social bs.bsky.social @zsusswein.bsky.social
If I can find someone I think there is a nice little project digging into these packages to see how they compare to methods more commonly used in epi
We have a converter to chainladder here: baselinenowcast.epinowcast.org/reference/as...
we should really have an issue to make one for reviser!
Thanks for the tag. I am actually writing a nowcasting overview paper for epis as we speak and have been writing about the various packages in economics and how they compare.
There is also chainladder: mages.github.io/ChainLadder/
which has a stan sibling somewhere.
GPJax 0.14.0rc1 is out!
Together with @theorashid.bsky.social, we've been able to remove 2200 lines of code whilst deepening our connection to the scientific JAX ecosystem, and streamlined our Numpyro integration to allow for bigger Bayesian models to be built.
Thank you for all the emails so far!
I'm back with another quarto extension, now presenting quarto-timelines!
Customizable timelines in html and revealjs slides, different layout and fragment support included
emilhvitfeldt.com/post/quarto-...
#quarto
Please and thank you!
I find this sort of stuff very hard to track as so often it doesn't show up in citations/isn't really public in the first place.
If can't flag on here and have something, an email would be much appreciated.
For reasons that are certainly unrelated to grant applications, does anyone know of (ideally with some evidence) uses of tools I have contributed to (scoringutils, epinow2, epinowcast, epidist, primarycensored, episoon, epinow, censoreddistributions ...) from different parts of the world?
(across language ports as well as within language as they are frustrated having to depend on others (fair)).
Seeing a lot of LLM induced "I don't need dependencies I will just port everything using AI" from more senior people who I think haven't historically had time to code but are now back using their LLM bots. Interesting to see how it goes.
Working through the materials for this 10 years or so ago was what made me want to work in real-time modelling of infectious diseases. The port to Julia has made it much more compelling in my opinion!
Really pleased that the model fitting and inference for infectious disease modelling short course is running again this year (7-10th September) at #LSHTM and is now all in #julialang.
www.lshtm.ac.uk/study/course...
More details about the Bayesian Workflow book and case studies now available on the book web site avehtari.github.io/Bayesian-Wor... (but you still need to wait a bit for the book)
New post on the @ropensci.org blog, "A Better R Programming Experience Thanks to Tree-sitter"
Edited by @etiennebacher.bsky.social, @davisvaughan.bsky.social, Steffi LaZerte
ropensci.org/blog/2026/04...
#RStats
If you care about the accuracy of the models, you should care about the accuracy of the language used when talking about the models.
I blogged and will keep adding more to this blog post series
I keep being impressed by the quality of @gregfolkers.bsky.social's monitoring of scientific articles, preprints, and news articles on infectious diseases -- One of the best accounts to follow to stay up-to-date!
No formal #epinowcast seminar this week but instead a show and tell plus open discussion on future community directions (3pm UK time today!):
www.epinowcast.org/seminars/202...
I'm looking for a post-doc to help organize Bayesian Data Analysis course avehtari.github.io/BDA_course_A... (200 students) and to do research on Bayesian workflow users.aalto.fi/~ave/publica... at Aalto www.aalto.fi/en, Finland. Background in Bayes needed. Up to five year contract possible.
Outbreak response and pandemic policy are difficult. Mainly in the sense that we've had several important ones since COVID (MPox, H5N1, now Men B) and any "lessons" from COVID are secondary to the harms from cuts to UKHSA and wider public health that no-one seems to care about.
Really pleased with how this one has come out. We take common output formats for household-based field epi studies, see how much structure we can impute from them, and make the case for reporting case and contact numbers stratified by household size.
