Anyone ever use the geometric Poisson distribution? If I've implemented it correctly in Stan, it's giving me better fits than negative binomial according to PSIS-LOO-CV with a big count occupancy model.

What's the best way to get started with a local LLM? Just got a new M5 Macbook (64 GB).

Hidden Markov movement models are actually kind of annoying as (1) discretising a continuous process isn’t even all that realistic or necessarily principled and (2) it comes at high computational expense.

What are some of the flagship long term vaccine health studies (i.e. investigating causal links to autism) with open data?

Yes.

Turns out continuous time hidden Markov model-step selection functions (HMM-SSF) with unequal sampling intervals sample like a dream in @mc-stan.org🔥

occARU now supports multiseason occupancy. The package (1) detects which sites were sampled in each season and then (2) fits a continuous time dynamic occupancy model for repeatedly sampled sites. 1/n mhollanders.github.io/occARU/artic...

occARU now supports sites within regions, e.g., multiple arrays of camera traps or microphones. Random (zero-sum) region-level intercepts are estimated for both occupancy and detection, and spatial site effects are computed per-region, which is much faster! 1/2 mhollanders.github.io/occARU/

How do you guys deal with journals requesting revised papers with Track Changes? I don't touch Word and exclusively use Quarto.

Addressing co-author comments on your responses to reviewer comments on a manuscript you (by now) despise is probably my least favourite part of academia.

occARU: Bayesian (multispecies) occupancy models for ARU data using @mc-stan.org. Automated recording units (ARUs) like camera traps produce rich time series which warrant going beyond occupancy and focusing on detection rates. 1/n mhollanders.github.io/occARU/

New pre-print! I cover a range of open capture-recapture models (single survey/robust design, multistate/multievent, in (Cormack-)Jolly-Seber variants) in Stan, and provide efficient log likelihood functions. I also introduce a method to account for unequal survey intervals in the entry process.

MC Stan@mc-stan.org · 4mo ago

"Efficient Bayesian implementations of capture-recapture models with Stan" by @matthollanders.bsky.social discourse.mc-stan.org/t/bayesian-c...

In (Gaussian) linear models, we usually put normal priors for regression coefficients. How do we feel about putting gamma(a, a) priors in something like Poisson regression? Using rates for rates seems nice; the geometric means are still 1, meaning the priors are centered on no multiplicative effects

I can’t stand that online journals can’t format equations properly.

What do people think of this? ai-2027.com

AI 2027

A research-backed AI scenario forecast.

ai-2027.com

Hey #stats people, what do we think about interpreting coefficients in cloglog binomial regressions? Since everyone hates odds ratios I wonder if interpreting hazard ratios is easier.

Why do we peer review for free?

I need some help with parameterising a latent simplex, so a set of latent probabilities that sum to 1. Can anyone have a yarn?

Hazard rates FTW 🎉

Quantecol@quantecol.bsky.social · last yr.

By default, mark-recapture models parameterise the survival process with survival probabilities. This is unfortunate, because logit-linear survival functions aren't time invariant, and you can't just exponentiate multistate transition probabilities like you can survival. Instead, we should... 1/2

brms is one of the most amazing R packages ever. Nevertheless I can't get myself to use it for my own work because nothing beats coding up your own models where you have ultimate control and insight into what's going on. There remains a "black box" element unless you are deeply familiar with it.

Working on a multisite/multistate/robust design Jolly-Seber model in Stan where I need to marginalise (1) the entry occasion of each individual and (2) the sub-area within each site where each individual enters. 😮‍💨

We just updated our HMM blog by applying the backward sampling algorithm to recover the posterior distributions of latent states. Check it out! quantecol.com.au/blog/hmm-in-...

Ecological Modeling in Stan

quantecol.com.au

Hey movement ecology people, are the models implemented in the #ctmm R package essentially just GPs fit to location data?

Launched a bsky page for the ecological stats consultancy, give us a follow!

Quantecol@quantecol.bsky.social · last yr.

Quantecol is on Bluesky! We are a statistical consultancy specialised in (but not limited to) quantitative ecology. We also blog about ecological statistics, with our latest post being about fitting various ecological models in Stan. quantecol.com.au/blog/margina...

Those are good looking plots.

Vincent Arel-Bundock@vincentab.bsky.social · last yr.

{tinyplot} 0.3.0 is out! 🚨 It's a lightweight #Rstats 📦 to draw beautiful and complex plots, using an ultra-simple and concise syntax. This is a massive release! @gmcd.bsky.social @zeileis.org and I worked hard to add tons of new themes and plot types. Check it out! grantmcdermott.com/tinyplot/

Another video in my series on applied time series and forecasting with the {mvgam} #rstats 📦. This one introduces State Space hierarchical GAMs and GPs for tackling multivariate series youtu.be/2POK_FVwCHk?...

Time series in R and Stan using the mvgam package: hierarchical GAMs

YouTube video by Nicholas Clark

youtu.be

New blog post! I show how to marginalise discrete variables from Bayesian ecological models in Stan and how to recover their posterior distributions after estimation. I cover occupancy and N-mixture models using multiple parameterisations and model types. quantecol.com.au/blog/margina...

Quantecol - Marginalisation

quantecol.com.au

I've been looking at different parameterisations of multi-season occupancy models, and it seems that dynamic (with colonisation and emigration) and auto-logistic models aren't all that different. Auto-logistic might be preferred with \alpha being a sort of average log odds. @masonfidino.bsky.social

I’ve been enjoying a restful Christmas break, and one of the highlights was this python selecting a carefully placed tin stack to lay her eggs.