#statsodon — Public Fediverse posts

also thanks to @gaborcsardi @maelle for rhub v2, which makes it easy to configure GitHub Actions for crossplatform builds on Linux, macOS and Windows

https://CRAN.R-project.org/package=rhub

#rstats #statsodon #github #jira

I mostly followed the 2nd edition of "R Packages" by @hadleywickham @jennybryan with a few minor tweaks to account for the differences between GitHub and Bitbucket/Jira

https://r-pkgs.org/release.html

I highly recommend this book for anyone who is writing or maintaining R code! #rstats #statsodon

New blog post! Seven (7!) new tidyexplain-esque animations showing how {dplyr}'s mutate(), summarize(), group_by(), and ungroup() all work together #rstats #statsodon https://www.andrewheiss.com/blog/2024/04/04/group_by-summarize-ungroup-animations/

New blog post! Have you (like me!) wondered what the ATT means in causal inference and how it's different from average treatment effects (ATE)? I use #rstats to explore why we care about the ATE, ATT, and ATU and show how to calculate them with observational data! https://www.andrewheiss.com/blog/2024/03/21/demystifying-ate-att-atu/ #statsodon

This paper by @nickchk (https://doi.org/10.1080/1350178X.2022.2088085 ; ungated here: https: //ftp.cs.ucla.edu/pub/stat_ser/huntington-klein-jem-june2022.pdf) is the best, most accessible introduction and explanation of how DAGs can be useful for causal inference for people more familiar with potential outcomes and econometrics-style approaches #statsodon #CausalInference

It's DAG day in class today and I *think* figured out a way to animatedly demonstrate collider bias (at least the selection bias version of it) #CausalInference #Statsodon

Many clinical studies evaluate the benefit of a treatment based on both survival and other continuous/ordinal clinical outcomes, such as quality of life scores. In these studies, when subjects die before the follow-up assessment, the clinical outcomes become undefined and are truncated by death. Treating outcomes as “missing” or “censored” due to death can be misleading for treatment effect evaluation.

#statsodon #StatisticsInMedicine

https://onlinelibrary.wiley.com/doi/10.1002/sim.9922?af=R

Interesting reading: Kernel Cox partially linear regression.

Building kernel Cox proportional hazards semi-parametric model and regularized garrotized kernel machine (RegGKM) method to account wide heterogeneity in cancer patients’ survival due to molecular profiles. #stats #StatisticsInMedicine #statsodon

https://onlinelibrary.wiley.com/doi/10.1002/sim.9938?af=R

“Split sample validation can require up to 20,000 observations to perform well enough. Otherwise the results may depend dramatically on the luck of the split. That’s why 100 repeats of 10-fold cross-validation or several hundred bootstrap resamples provide better estimates of likely model performance in most cases.”

Frank Harrel’s answer is a must-read.

#statistics #stats #statsodon

https://stats.stackexchange.com/questions/627752/too-good-to-be-true-ridge-prediction

#R/Pharma is organizing a bunch of amazing workshop for its 2023 conference. They are online and free and they will take place in October. In case you're interested: https://rinpharma.com/workshop/2023conference/

#rstats #Statsodon #statistics

Update! The actual reason I had to figure out this distribution is because the ordered beta model I was using used it to define priors for the values *between* the Dirichlet columns. The post now shows how to work with those cutpoints #rstats #statsodon #Bayesian https://www.andrewheiss.com/blog/2023/09/18/understanding-dirichlet-beta-intuition/#bonus-later-addition-boundaries-between-categories

New post! Have you (like me) been confused/intimidated by Dirichlet distributions? Here's a basic, visual-heavy, intuition-heavy guide to the Dirichlet distribution. It's just a Beta distribution, but ~fancy~! #rstats #statsodon https://www.andrewheiss.com/blog/2023/09/18/understanding-dirichlet-beta-intuition/

yessssss this brms bayesian model for a conjoint survey experiment took 3 hours to fit, estimated nearly 40,000 unique parameters, takes up 4 GB of space, and is most definitely way overkill, but IT CONVERGED AND WORKS GLORIOUSLY #rstats #bayesian #statsodon

Also, in the course of adding DOIs to past posts, I updated my big ol' guide to different flavors of marginal effects to use {marginaleffects}'s newer slopes(), predictions(), and comparisons() functions https://www.andrewheiss.com/blog/2022/05/20/marginalia/ #rstats #statsodon

Q about #bayesian stuff: I'm finding the probability of direction (proportion of posterior that's >0 or <0), but I never know how to report these, since sometimes they're positive and sometimes they're negative. My current solution is to use a column for each direction—is there a better way tho? #statsodon

Check out this new ultimate guide to multilevel/hierarchical multinomial conjoint analysis with #rstats and {brms}, including how to find both marketing-style predicted market shares *and* polisci-style causal effects *across individual covariates* #bayesian #statsodon

https://www.andrewheiss.com/blog/2023/08/12/conjoint-multilevel-multinomial-guide/

heck yes full-luxury multilevel multinomial models with brms #rstats #bayesian #statsodon

Probably super naive multilevel modeling question that I'm way overthinking (https://discourse.mc-stan.org/t/does-this-hierarchical-brms-formula-correctly-link-covariates-across-the-multiple-nested-levels/32410):

I have data like this, with variables measured at different levels. In raw Stan, this can be estimated w/2 diff datasets.

Does this brms formula do the same thing? #rstats #bayesian #statsodon

Here it is! The ultimate practical guide to Bayesian and frequentist conjoint data analysis with #rstats and {brms} and {marginaleffects}, including how to distinguish between marginal effects and marginal means + work with subgroups! #statsodon https://www.andrewheiss.com/blog/2023/07/25/conjoint-bayesian-frequentist-guide/

OLS Strikes Back
#academia #statsodon #SAG
ht @WeedenKim

throwing some do() operators into this conjoint blog post so you know it's serious causal inference #CausalInference #statsodon

regular PSA for those working with survey data and Likert scales:

It's "lick-ert" not "like-ert"

https://en.wikipedia.org/wiki/Likert_scale#Pronunciation #statsodon

New blog post! Here's a guide to calculating the differences between categorical proportions in a principled, #bayesian way with #rstats, #mcmcstan, and {brms}, including fancy things like mosaic plots (with {ggmosaic} and striped fills (with {ggpattern}) https://www.andrewheiss.com/blog/2023/05/15/fancy-bayes-diffs-props/ #statsodon

Using binomial(link = "identity") in a #brms model to get differences in proportions between different categories, and it feels wrong but it works really well and is actually kinda neat 🤷‍♂️ #rstats #bayesian #statsodon

(Here's how to do it with tidybayes and marginaleffects: https://gist.github.com/andrewheiss/a1f2837bd4e5ba40280e81df1efbe720)

random walk, n., the route through an open plan office taken by a #statistician whilst waiting for their #MCMC chains to converge (hopefully). Expected length of route may depend on factors such as hardware specifications, informativeness of #priors, and whether there's a good coffee machine nearby. #statsodon #Bayesian #iamworking @pymc

random walk, n., the route through an open plan office taken by a #statistician whilst waiting for their #MCMC chains to converge (hopefully). Expected length of route may depend on factors such as hardware specifications, informativeness of #priors, and whether there's a good coffee machine nearby. #statsodon #Bayesian #iamworking @pymc

random walk, n., the route through an open plan office taken by a #statistician whilst waiting for their #MCMC chains to converge (hopefully). Expected length of route may depend on factors such as hardware specifications, informativeness of #priors, and whether there's a good coffee machine nearby. #statsodon #Bayesian #iamworking @pymc

random walk, n., the route through an open plan office taken by a #statistician whilst waiting for their #MCMC chains to converge (hopefully). Expected length of route may depend on factors such as hardware specifications, informativeness of #priors, and whether there's a good coffee machine nearby. #statsodon #Bayesian #iamworking @pymc

ahhh this new #bayesian ordered beta regression model family (package: https://github.com/saudiwin/ordbetareg_pack; paper: https://doi.org/10.1017/pan.2022.20) is so so neat! I have an outcome bounded at 1 and 32, and the model successfully predicts discrete 1s and 32s, as well as a continuous range in between! #rstats #brms #statsodon

Talking about DAG colliders in my program evaluation class tonight, so back to my good ol' standby examples of niceness → appearance + race → police use of force #statsodon #CausalInference

Love this consistency! #rstats #statsodon

I love this new version of @vincentab ‘s {marginaleffects}! The whole workflow has been streamlined so you can get predictions with predictions(), contrasts/comparisons with comparisons(), and partial derivatives with slopes() + single values with avg_*() + plots with plot_*()! #rstats #statsodon

https://fosstodon.org/@vincentab/109789722872680890

The 8th iteration of my #ProgramEvaluation and #CausalInference course is up and live at https://evalsp23.classes.andrewheiss.com/ !

It covers basic econometrics and DAGs, all with #rstats, and it's mostly asynchronous with dozens of hours of videos, and the whole thing is Creative Commons-licensed, so do whatever you want with it! #epitwitter #EconTwitter #statsodon

This latest episode of the Casual Inference podcast on instrumental variables is fantastic and it’s neat to hear about IV from a non-econometrics perspective https://casualinfer.libsyn.com/website/instrumental-variables-with-maria-glymour-season-4-episode-5

and my post on conditional and marginal effects makes a surprise appearance at the beginning lol https://www.andrewheiss.com/blog/2022/11/29/conditional-marginal-marginaleffects/

#statsodon #causaltwitter