Real-time infectious disease modelling. Developing methods for outbreak response, surveillance, and pandemic preparedness. https://xitter.cf/t.co/qIESZxy6rY
Bristol, England
Joined October 2015
- Tweets5K
- Following817
- Followers2.7K
- Likes11.2K
Pinned Tweet
We started this with the aim to provide both decision-makers and the general public with real-time information on how the epidemic was progressing
So was it useful? Hard to tell.
epiforecasts.io/posts/2022-0…
Sam Abbott (@[email protected]) retweeted
For this weeks #epinowcast seminar (Weds at 3pm UK time (10 am ET)) we have @foxandtheflu.bsky.social
talking about Optimizing disease outbreak forecast ensembles and the use of microHubs
epinowcast.org/seminars/2024…
Modelling epidemiological delays being useful
cdc.gov/mmwr/volumes/73/wr/m…
Some nice work from folks at the CFA CDC and partners
Really enjoyed the first IDM conference in Thailand!
Some great talks and a great way to meet a whole lot of new folks at once.
We had a massive turn out at our understanding and evaluating forecasting workshop which was great and left me with some things to think about on the flight home. If you missed it the material is here nfidd.github.io/ueifid/ or come along to our course next June in Stockholm.
I also enjoyed talking through the last five years of thinking about situational awareness. Slides are here with lots of links out.
samabbott.co.uk/presentation…
I had some good feedback that people enjoyed the focus on mistakes so planning to keep making plenty more.
Sam Abbott (@[email protected]) retweeted
Seminar today at 3pm UK time with Matthew Shin talking about:
Quantifying regularity of COVID-19 epidemic waves in France to assess predictability of future waves and improve forecasting
epinowcast.org/seminars/2024…
primarycensored #rstats 📦 is now on CRAN and stable.
primarycensored.epinowcast.o…
If you want to use or estimate delay (think incubation periods, reporting delays, etc.) distributions in epi it might be worth checking out.
A #julia version is in development.
An absolutely huge amount of work from @nikosbosse on this starting before his PhD and continuing completely through it to now as a side project to his full time ai startup gig.
Great job handling inputs from lots of people whilst learning on the job.
Happy to share that we FINALLY published version 2.0.0 of scoringutils on CRAN.
scoringutils is an R package designed to facilitate the evaluation of forecasts. It operates on data.table, is very slick and has a lot of visualisations and extra features.
epiforecasts.io/scoringutils…
This sort of work is really badly incentivised in academia especially the long term maintenance. Very little change appears to be happening on that. It’s also not really possible in places like public health agencies despite ongoing efforts to change this.
Sam Abbott (@[email protected]) retweeted
Happy to share that we FINALLY published version 2.0.0 of scoringutils on CRAN.
scoringutils is an R package designed to facilitate the evaluation of forecasts. It operates on data.table, is very slick and has a lot of visualisations and extra features.
epiforecasts.io/scoringutils…
Woo! Really good reviews on this have really improved it I think. Check it out if estimating or working with delay distributions in epi.
Publication day for my first perspective piece 🎉 Lucky to have worked with such great collaborators on this project 🤩 Read it here: journals.plos.org/ploscompbi… @institutpasteur
If you want to implement some of the ideas from this work primarycensored just hit CRAN (Julia to follow soon)
primarycensored.epinowcast.o…
This will all soon be supported epidist to allow even more efficient, flexible, and accurate delay estimation at scale
epidist.epinowcast.org/artic…
Woke up today and realised I had a big backlog of PRs to take a look at.
Made me reflect and I've realised over the last few years that I've nearly moved away from writing code as my main form of software dev... feels weird!
Okay, I hear you Sam being Bayesian is cool but what if like I just don't want to be right now?
primarycensored.epinowcast.o…
(fit primary censored and truncated distributions using fitdistrplus)
I can't think of a better PhD supervisor. Lots of impactful work backed up with solid theory from Bracher et al.
The interface between nowcasting and forecasting contains lots of interesting untapped ideas and unlike many parts of IDD a lot of the work is tractable.
PhD position (Karlsruhe, Germany)
Nowcasting and Short-Term Forecasting of Infectious Diseases
with @johannesbracher
at Karlsruhe Institute of Technology
More details: iddjobs.org/jobs/phd-candida…
Sam Abbott (@[email protected]) retweeted
Reminder I’m still looking for a postdoc to join the Fox lab! Starting to review applicants this and next week, but email me if you are interested and can’t get to the application in time.
