Adam Fleischhacker @preposterior@qoto.org

Made an introductory 📕(draft) about using Python for Bayesian Inference and unifying narrative, math, and code. People seem to find it helpful. Check it out. Feedback encouraged.

https://persuasivepython.com

#DataScience #Python #bayes #Stats #probabilisticprogramming

@Cmastication @brohrer 100% Truth. Now I just need some type of inspirational proposition for when you put alot of effort into the "trying it out" only to find out it doesn't work as well as hoped. The (un)productive Struggle is real.

Know anyone who wants to teach #rstats and #Bayesian statistics out of a business school? Please make them aware of the second edition of my textbook. Students love it and it is a great entry point into data science. https://www.causact.com/

For better or worse, I self-illustrate my textbooks despite having zero artistic ability. On my latest edition (coming out next week), I used NightCafe AI image generation to get all artsy with my book cover. I hope you like the AI art... here is a pre- and post-ai book cover comparison.

@andrew have you tried DiagrammeR in #rstats with sugiyama layout? would work well for the plots pictured. no need to position everything manually. tikz is both beautiful and painful.

@andreashandel thanks! feel free to reach out with questions, suggestions, or any feedback.

Updates to supporting textbook coming soon. But dag_numpyro() is now a drop-in replacement for dag_greta(). Non-updated book here: www.causact.com

Simple #rstats install - works on ALL platforms (see readme for more). Just four steps. Please test this #numpyro version before I release to CRAN.🙏

install.packages("remotes")
remotes::install_github("flyaflya/causact")
library(causact)
install_causact_deps()

The {causact}📦 is bringing #numpyro speed to #rstats. Learning computational #Bayesian inference has never been easier.

🐣Install with ease
📊Build visual models with DAGs
🚅Automate Inference at #numpyro speed

#numpyro now available via github:
https://github.com/flyaflya/causact

Bayesian Additive Regression Trees (BART) seem like a robust way to do causal inference; easily-queried black boxes with built-in uncertainty quantification of causal effects. Anyone using BART in practice?

@danielyuksek you have my two sites: causact.com and persuasive python.com. You can also check out the free videos and resources by @rlmcelreath at https://xcelab.net/rm/statistical-rethinking/. For masterful foundational treatment see @betanalpha work at https://betanalpha.github.io/writing/. Hope that helps.

Check out Sal's outlook on AI in education. A segment where a student engages the iconic Jay Gatsby is great. I'm presently immersed in Atlas Shrugged and appreciate being able to request "a book review of Atlas Shrugged written by AOC."
https://youtu.be/hJP5GqnTrNo

I've been making progress on my #Bayesian inference in #python book draft. Latest addition is the multi-level modelling chapter. Check it out and then post your feedback here. Thanks! #numpyro

BOOK Link: https://www.persuasivepython.com/13-multilevel

@davidmanheim what a wildly fun read this is! hysterical, I wonder how much truth is in these words?

Most enjoyable article-writing experience ever thx to gr8 co-authors🔥.

We find how to value integer-based data prior to seeing it; a magic trick with quadratic loss being hidden secret to get closed-form🤫. Link to article in advance of publication:

https://pubsonline.informs.org/doi/10.1287/deca.2022.0462#.Y7Lvk3xoofQ.twitter