iiinova — where your digital self keeps evolving

Stats For Bios @statsforbios.bsky.social · 05/10/2026

New preprint from my student Carolina Molina-Tabares. Different to other preprints I post! She evaluated a portable qPCR workflow for rapid on-site detection of Ae. aegypti under field conditions. Compared to benchtop Roche LightCycler 96. Tested in Kenya. www.researchsquare.com/article/rs-1...

Photo and diagram of the field setup. A portable qPCR machine, run on a motorcycle battery, plugged into a laptop.

000

Stats For Bios @statsforbios.bsky.social · 17/09/2026

Had a great week teaching Geospatial Models for Public Health at Tanzanian National Bureau of Statistics. Been supporting Janine Illian teaching inlabru and teaching disagapp for disaggregation regression. inlabru-courses.github.io/Tanzania_202... Now get the train back to Dar es Salaam

View from train from Dar es Salaam to Dodoma.

130

Stats For Bios @statsforbios.bsky.social · 11/09/2026

Preprint shinyapp by @simonsmart.bsky.social here. User-friendly, reproducible software for high-resolution predictions from aggregate polygon case counts (or other data). Once you're done, click a button to get a Rmd file out. #RStats #DiseaseMapping #malaria www.medrxiv.org/content/10.6...

164

Stats For Bios @statsforbios.bsky.social · 01/09/2026

Histogram of number of pages for your viewing pleasure. The 2000 page one is an error, but the other big ones are fairly accurate.

Histogram of number of pages in each book. Quite strong skew. Mode at about 400. A few greater than 1000 and one at around 2000.

010

Stats For Bios @statsforbios.bsky.social · 13/08/2026

Album of the week. A black metal (blackgaze really) album about Bill the Pony. I mean look at that cover.

A cute cartoony illustration of Bill the Pony.

050

Stats For Bios @statsforbios.bsky.social · 07/08/2026

Estimating the effect of short term (hourly) air pollution is difficult but that's what we tried to do here. We used data that, unusually, has admission time. We found significant associations between NO2 and asthma admission at 3-6 hour lags but not other lags. bmjopen.bmj.com/content/16/8...

The relative risk (RR) of emergency hospital admissions associated with mean nitrogen dioxide at the specific lags from 0 to 24 hours. Note: the red line shows the average RR of emergency hospital admissions for acute asthma, and the error bars are the 95% CIs.

Lags between 0 and 2 hours are insignificant, with large uncertainties at 0 hours. Lags between 3 and 6 hours are significantly positive, with CIs around 1,005 and 1.04. After 6 hours, all CIs cover zero again.

101

Stats For Bios @statsforbios.bsky.social · 05/06/2026

New preprint from my student Nidhi Shukla. She developed an easy to use method for disaggregating disease risk over time. If you only have daily disease counts, can you estimate the effect of hourly exposures? www.medrxiv.org/content/10.6...

Two curves representing estimated relationships between temperature and respiratory mortality risk. One is using daily disease counts and daily temperatures. The other uses weekly disease counts and daily temperatures. Both curves are qualitatively similar and CIs largely overlap.

011

Stats For Bios @statsforbios.bsky.social · 18/05/2026

Great work from @clareecenevill.bsky.social and others. "Developing Public‐Friendly Visualisations to Improve PPIE Glossaries for Stats" Really interesting to see how the visualisations progressed, often deviating strongly from initial sketch. onlinelibrary.wiley.com/doi/10.1111/...

010

Stats For Bios @statsforbios.bsky.social · 11/05/2026

Fuck yeah! Bullet game with no mistakes.

Screen shot from chess.com app showing the review of my game with no blunders, misses or mistakes.

020

Stats For Bios @statsforbios.bsky.social · 03/05/2026

Don't mean to be rude, but this is an awful way to present this data. #DataViz www.reddit.com/r/dataisbeau...

Two maps. Lines come from various countries and link to a horizontal bar graph of house price to income ratio.

220

Stats For Bios @statsforbios.bsky.social · 24/04/2026

Just an interesting graph I found while discussing the difference between my dinosaur books and my nephew's books, with my mum.

A line plot from the year 1840 to 2000. One line is the total number of dinosaur genera. The other is the number of dinosaur genera that are still considered valid. Both increase very rapidly from the 80s.

1151

Stats For Bios @statsforbios.bsky.social · 20/04/2026

Nice design detail from a poster by @hadiqa-tahir.bsky.social. I really like how the figure crosses intro and methods with "existing methods" in the intro and "our new method" in the methods. Not quite #DataViz, but probably overlap.

151

Stats For Bios @statsforbios.bsky.social · 17/04/2026

And last one for now. I was presenting this poster on behalf of Amber Vayani (hence no person in the photo). A big applied project mapping air conditioning in lmics. Impacts on mosquito borne disease and extreme heat related health outcomes.

A poster with a map of predicted air conditioning ownership

010

Stats For Bios @statsforbios.bsky.social · 17/04/2026

More work from Leicester colleagues at "Environmental Modelling meets Infectious Diseases". Sylvain Matingou presenting methods for including local movement into machine learning models for malaria mapping. #EpiSky #MachineLearning

050

Stats For Bios @statsforbios.bsky.social · 16/04/2026

Hadiqa Tahir presenting her work on directly including movements models into environmental epi regressions. #EpiSky #StatsSky @hadiqa-tahir.bsky.social

022

Stats For Bios @statsforbios.bsky.social · 16/04/2026

Poster describing disagapp by @simonsmart.bsky.social. App: disagapp.le.ac.uk Code: github.com/simon-smart8... #RStats #EpiSky

031

Stats For Bios @statsforbios.bsky.social · 16/04/2026

And now Samana Bano presenting on new methods/software for fitting spatial disaggregation regression model with data through time. Quite a few people have been asking for this! Watch this space for software. #StatsSky #EpiSky

010

Stats For Bios @statsforbios.bsky.social · 16/04/2026

Olukemi Olowofoyeku presenting new methods that combine disaggregation regression and MaxEnt style feature engineering. New method uses the covariates more, doing more "disaggregation". #StatsSky #EpiSky

100

Stats For Bios @statsforbios.bsky.social · 30/03/2026

Makes me so happy when my #pokemon colour palette #RStats package gets cited. onlinelibrary.wiley.com/doi/full/10....

"Colors for the plots were gathered from the pokepal palettetown package (Lucas 2016)."

020

Stats For Bios @statsforbios.bsky.social · 25/02/2026

I "invented" a "thing". For taking very rough, on the fly notes in a book. A folded corner and a couple of scores with my thumbnail points clearly enough to the bit I was interested in. I can often then remember what I thought about it. Obviously has limitations.

000

Stats For Bios @statsforbios.bsky.social · 22/01/2026

Biological species concept is all well and good. Then you start reading about Bengal cats and Savannah cats and it's clearly all nonsense. upload.wikimedia.org/wikipedia/co...

050

Stats For Bios @statsforbios.bsky.social · 19/12/2025

"the task of replacing a burnt-out light bulb may turn out to require moving a garbage bag; this may unexpectedly cause the spilling of a box of pills, which then forces the floor to be swept so that the dog won't eat the pills." Glad to see Hal's lightbulb is in Gödel, Escher, Bach

121

Stats For Bios @statsforbios.bsky.social · 20/11/2025

Sorry this has all got me on a quick nostalgia kick. This is the venue I learned to love live music in (the Freebutt). It got closed down because they built residential houses next to it, people moved in and didn't like the noise. And now it's just nothing. What an absolute waste.

A run-down, closed-up venue. Looks pub size.

010

Stats For Bios @statsforbios.bsky.social · 09/11/2025

Anyone know this film? It's on in the bar and appears to be completely bonkers. But I have no idea what it is

100

Stats For Bios @statsforbios.bsky.social · 07/11/2025

If anyone's going to ASTMH a day early and likes punk, pretty fun looking gig on Saturday. I'll be there if my flights behave. files.elfsightcdn.com/eafe4a4d-343...

Gig poster. Bovine sex club venu.e Decade, wolfcharge absolutely and manic. 9pm, nov 8th, $10.

000

Stats For Bios @statsforbios.bsky.social · 20/10/2025

New paper, great work by Hannah Worboys. Feasibility study of the P-STEP app. "At 12 weeks, the mean (sd) System Usability Score was 61.68 (22.9)". And lots of useful prelim data for larger studies. #EpiSky #EnvironmentalEpi #copd #asthma

Four histograms. a) Usage per week (mostly between 1 and 6 times a week), b) likely to recommend (mostly likely), c) app rating (mostly 4-5 stars) and d) attitudes to future use (almost 50/50 on future use).

010

Stats For Bios @statsforbios.bsky.social · 29/09/2025

Wait, stata costs twice as much if you want to run it using 2 cores? And even more for 4 cores? Lol.

A table of stata prices. Standard is $94. "MP 2 core" is $275.

020

Stats For Bios @statsforbios.bsky.social · 25/07/2025

Patent infringment from Wallace (& Gromit) coming any day now.

1110

Stats For Bios @statsforbios.bsky.social · 23/07/2025

Dawww. A baby bee. Please don't a) tell me it's not a baby bee or b) tell me it's some horrible invasive. Enough of a rollercoaster day already.

100

Stats For Bios @statsforbios.bsky.social · 17/07/2025

Got case data aggregated to admin areas? Got important covariates in high-resolution rasters/grids? We are running a free online training workshop on user-friendly, interactive software for this use-case. Disaggregation regression via the shinyapp disagapp. github.com/simon-smart8... #malaria

Screenshot of disagapp. Shows malaria data from madagascar aggregated to admin regions on an interactive map. A nicely designed user interface provides modelling capabilities.

193

Stats For Bios @statsforbios.bsky.social · 11/07/2025

You often have to choose the number of clusters. If clusters are clearly arbitrary and just a modelling tool, fine, look at some plots and choose. But when the main result of the paper is "there are 4 clusters and this reflects reality and here's our description of the clusters". It's so weak.

6 plots showing metrics of a model with different numbers of components, 1-10. Log likelihood, AIC, BIC and others are shown. They all very smoothly improve with number of clusters. No clear cut jumps or elbows.

4110

Stats For Bios @statsforbios.bsky.social · 02/07/2025

I so often find dual axis graphs confusing while I sort out what's what. Having it spelled out like this makes a huge difference. #DataViz

A dual axis graph. Two lines showing two variables changing through time. Text, next to each line, says what the variable is and which y axis it corresponds to.

160

Stats For Bios @statsforbios.bsky.social · 23/06/2025

Our new paper led by Jenicca Poongavanan. Predicting #dengue in Africa using disaggregation regression (bottom right in the inset image) and comparing to other methods. www.biorxiv.org/content/10.1...

A compound figure. The main image is of Africa with colour showing the number of models that agree in each pixel. 4 models agree in areas of west africa. 3 models agree in a large area should of the Sahara. The inset has four binary prediction maps for the different models.

144

Stats For Bios @statsforbios.bsky.social · 11/06/2025

New paper led by @esnightingale.bsky.social. Delay to diagnosis is important for visceral leishmaniasis control. Longer delay -> more transmission. Can we understand these delays, factors correlated with them and spatial distribution. bmcglobalpublichealth.biomedcentral.com/articles/10....

3 maps of Bihar India. Map of estimated delay is diagnosis. Posterior probability that the delay is greater than 30 days. And the raw block mean delay.

010

Stats For Bios @statsforbios.bsky.social · 07/06/2025

Messing around with my last.fm scrobble data again. Each year, who were my 5 favourite new (to me) bands? #RStats

A screen shot of a table from R. Columns are artist, scrobbles and year.
Excerpt:
87 Andy B & The World 282 2022
88 Celeste 158 2022
89 Israel Vibration 156 2022
90 Astronoid 131 2022

250

Stats For Bios @statsforbios.bsky.social · 10/05/2025

Does America have anything roughly equivalent to a potato scone? I can imagine they'd be very popular. en.wikipedia.org/wiki/Tattie_...

A potato scone. It is a fried pancake type thing. Golden brown, glistening slightly from the butter, served with a mega crunchie looking thing that I can't remember of the name of and a fried tomato. The adjacent fried bacon and black pudding is implied but not directly shown in the image. Availability of brown sauce is open to interpretation.

110

Stats For Bios @statsforbios.bsky.social · 09/05/2025

New work by my student Flo Goemans (and others). No change between sexes in accuracy of polygenic risk scores for idiopathic pulmonary fibrosis. Surprising given higher prevalence in males. publications.ersnet.org/content/erjo...

AUC curves for prediction of idiopathic pulmonary fibrosis using polygenic risk scores. Separate curves for male, female and overall but all three curves are very similar.

120

Stats For Bios @statsforbios.bsky.social · 30/04/2025

If I'm reading this right, it's dual axis, on totally different scales (% Vs millions) but the numbers just happen to line up exactly. Pretty crazy and more confusing than if they were different. #DataViz

Horizontal bar plot of ww2 deaths by country. Has deaths in millions and deaths as a percent of the country. Top and bottom axis have identical numbers on the axis ticks.

000

Stats For Bios @statsforbios.bsky.social · 11/04/2025

Thanks #EuroCIM2025! Great conference. Now coding up my first ipw for malaria inference on my way home. #StatsSky #CausalInference #CasualInference

A photo of a laptop with maps of environmental variables in Madagascar and some R code. In a bar (in Brussels) with quite a nice plant.

030

Stats For Bios @statsforbios.bsky.social · 26/02/2025

Not sure I've ever used geom_smooth() like this to make essentially a calibration plot. Super easy. Also, hot damn my simple model is well calibrated. Predictive performance is bad, but calibration is gooood. # Rstats

ggplot2 scatter plot with predicted probability of diabetes on x axis and diabetes (0 or 1) on y axis. Geom smooth creates a very straight line that almost exactly matches the reference slope=1, intercept = 0 line.

110

Stats For Bios @statsforbios.bsky.social · 22/01/2025

Inelegantly answering my 9yr old niece's homework. #RStats

180

Stats For Bios @statsforbios.bsky.social · 25/11/2024

Painting social with Leicester biostats

040

Stats For Bios @statsforbios.bsky.social · 22/11/2024

Nice to see disaggregation regression used in more areas. "Our findings show that poorer communities are more prone to selling cattle in response to precipitation shortages, a practice that can erode long-term resilience and deepen inequalities." www.researchsquare.com/article/rs-5...

Maps of cattle slaughter in South Africa. A) raw data aggregated to province level. B) Estimates of disaggregtted cattle slaughter at high resolution (16km pixels).

040

Stats For Bios @statsforbios.bsky.social · 21/11/2024

Hooray! Our paper (led by Emily Nightingale) is out. Inferring the regional distribution of Visceral Leishmaniasis incidence from data at different spatial scales. w/ @grahammedley.bsky.social and others. www.nature.com/articles/s43...

Panel plot. a) shows estimated coefficients from models with and without a spatial random field. NDVI is positive for both. LST is strong negative until you add a field. b) predicted rate of VL. High in the north of Bihar, India. c) The fitted random field. High in North Bihar as well.

1164