Miles Moore

Most people don’t know this but Abraham Lincoln’s close male friend who he shared a bed with actually designed the original Lincoln Bathroom

It's like LLMs in the past few months realized that little reprexes are cool, but people aren't realizing that and just going with the fake code. Can't wait for this to inevitably make it into published stuff.

Graphical summary of 10 quick tips to get started with Bayesian statistics and how they fit into a larger view of an analytical workflow with Bayesian models.

🚨🎉 New paper "Ten quick tips to get you started with Bayesian statistics", hope you'll like it 😇🤗

✍🏽 w/ Andy Royle, Marc Kéry and Chloé Nater

🔗 dx.plos.org/10.1371/jour...

@plos.org @cnrsecologie.bsky.social @cnrsoccitanieest.bsky.social @umontpellier.bsky.social

Prior Diagnostics and Sensitivity Analysis Provides functions for prior and likelihood sensitivity analysis in Bayesian models. Currently it implements methods to determine the sensitivity of the posterior to power-scaling perturbations of the...

priorsense 1.1.1, for prior and likelihood sensitivity analysis, is now in CRAN (n-kall.github.io/priorsense/)

There is now an easier way to select which priors are power-scaled using the new prior tags feature in brms (2.22.11+), which allows focusing the analysis on specific priors.

1/3

I've never met a good person who went out of their way to belittle waiters

A slide with the text “Predicting a missing variate for a fully observed covariate, however, is not always the relevant predictive task.”, a probabilistic graphical model for predicting a missing variate given a partially observed covariate, and a corresponding equation for the posterior prediction distribution.

On Wed Dec 10 join me to learn about what a regression model is, and what a regression model is not, while raising funds for World Central Kitchen and United Farm Workers. For details on how to register, sponsored registration possibilities, and more see betanalpha.github.io/courses/.

i'm sorry WHAT

Our new pre-print is up on BioRxiv Ecology - goal was to better estimate how trait evolution impacts population dynamics & community composition. We (me, postdoc Ruben Hermann) used a Joint Species Distribution Models (HMSC) fit to simulated & empirical eco-evo data.

www.biorxiv.org/content/10.1...

writing is thinking (code edition)

Just because an LLM can produce a report with various figures & charts doesn't mean it is good at statistics.

Because good statistics is not about producing code.

It's about deep knowledge of study design & conduct. In my opinion, 95% of all data science problems come from poor questions & design.

Automating inferential modeling with AI is absolutely a step in the wrong direction. Rstats, by way of abstracting away a lot of the dirty details of a model, has already created *some* bad statistical habits among scientists. Adding another layer btw scientist and their data is not a net benefit

Interested in simulating the kind of data that you might commonly find in evolutionary and ecological studies?

Then we have the R package for you - squidSim!!

Check our new preprint:
ecoevorxiv.org/repository/v...

This only happens to you once

German classified ad: 1 skeptisher Hamster zu verkaufen 20€ Art: Hamster Beschreibung: Er guckt einen skeptisch an, als würde man nichts richtig machen. Es macht mich wahnsinnig, ich kann diesen vorwurfsvollen Blick nicht länger ertragen. Sein Name ist Olaf.

Was reminded of this classic today - we are all Olaf when reading the scientific literature

Figure 5. Time series showing cumulative number of cases averted at each time caused by the intervention calculated using our method (single-world) and a standard method. Shaded regions denote 90% confidence intervals. Note that there is more variation in the middle of the epidemic, so it may seem as though the number of cases averted is large during those times. (Online version in colour.)

Are we doing simulations wrong? This paper convinced me we are. doi.org/10.1098/rstb... Usually we run 2 sets of "worlds" w and w-out intervention. Gives large uncertainties that include negative (harm) effects of interventions that are actually always positive (beneficial)!

Thank you! Very excited and grateful to be in a position to do so

I just submitted my first ever first author paper! Whoop! So cool

Betteridge's law of headlines? en.m.wikipedia.org/wiki/Betteri...

before and after today’s glacier collapse that buried 90% of blatten, switzerland

Tweet: Stats twitter: “All statistical methods fail horribly without telling you and you shouldn’t trust any research ever. fml.” Data science twitter: “Here are NINE SIMPLE WAYS to convert your data into ACTIONABLE INSIGHTS using POWERFUL AI. Number 3 will astound you!! I’m on a boat”

Among my favorites, from @cameronpat.bsky.social

Knud Andreassen Baade, Norwegian (1808-1879), Norges vestkyst (On the West Coast of Norway), 1852, oil on canvas, Historisches Museum Bamberg, Germany

Are you excited about the elegant philosophy of Bayesian inference, but struggling to see how it can be applied beyond the idealized examples in introductory texts and tutorials? Over the past few months I’ve released a series of demonstrative analysis that might help. 👇

GitHub - benjamin-rosenbaum/bayesian-intro: Introduction to Bayesian statistics Introduction to Bayesian statistics. Contribute to benjamin-rosenbaum/bayesian-intro development by creating an account on GitHub.

I completely re-worked my course "Introduction to Bayesian statistics with brms" and taught it for the 1st time this week. It is meant as a tutorial for ecologists, but should be general enough for other sciences as well
github.com/benjamin-ros...
#stats #Rstats #brms #Stan #Bayesian

I have been using Python a lot more recently and just use plotnine for visualization 99% of the time. Seaborn isn’t bad though!

Science Communication as a Moral Imperative We need to do a much better job of encouraging scientists to be stronger communicators, and share the wonders of science, and the important…

We need to see science as a social compact where it serves the larger world, not our own interests.

As a first step, scientists must communicate to the larger world much better.

I wrote an essay about this in 2016, which still holds today.

globalecoguy.org/science-comm...

Foundations and future directions for causal inference in ecological research, forthcoming w/ @katherinesiegel.bsky.social: tinyurl.com/4mm57zzd.
We wrote it based on our experience teaching causal inference to ecologists w/ some stuff we wish other papers had reviewed to share w/ the students
1/n

You know, your own imagination is far more wonderful Mister Rogers’ Neighborhood

Than any computer could ever be. You know why? Mister Rogers’ Neighborhood

You’re a living human being. And a computer is just a machine. Mister Rogers’ Neighborhood

Human beings are far more wonderful than machines. Mister Rogers’ Neighborhood

Mister Rogers never misses

New timeline, same problems, same solution

A Graphical Catalog of Threats to Validity: Linking Social... : Epidemiology lenging, but these skills are often glossed over in training. Campbell and Stanley’s framework for causal inference has been extraordinarily influential in social science training programs but has rec...

Threats to validity (with helpful DAGs)

journals.lww.com/epidem/fullt...

On The Tardiness Of Coworkers And How To Exploit It - Journal of Astrological Big Data Ecology In this groundbreaking research, we determine the optimum amount of time to show up late to a meeting by arbitrarily assigning numbers to things

A little late to the party here, but just read a fascinating (totally real) paper on applied gaussian processes. Titled "On the tardiness of coworkers and how to exploit it": jabde.com/2023/02/12/c...