Michael Sachs (@sachsmc) header
Michael Sachs (@sachsmc) profile picture
Sweden
Joined May 2009
Photo by Michael Sachs (@sachsmc)Photo by Michael Sachs (@sachsmc)Photo by Michael Sachs (@sachsmc)Photo by Michael Sachs (@sachsmc)Photo by Michael Sachs (@sachsmc)Photo by Michael Sachs (@sachsmc)Photo by Michael Sachs (@sachsmc)Photo by Michael Sachs (@sachsmc)Photo by Michael Sachs (@sachsmc)Photo by Michael Sachs (@sachsmc)
Around Copenhagen and curious about data analysis and/or #RStats? Come to the CopenhagenR meetup! On the 5th of November we host a presentation on 🚀Blazingly fast data manipulation using r-polars and rust from R🚀 Everyone is welcome! 😊 meetup.com/copenhagenr-user-…
2
10
614
Michael Sachs retweeted
excited to see the physical copy of the book; nervous about the potential errors. Comments are welcome.
I just uploaded the R code and datasets to Harvard Dataverse: doi.org/10.7910/DVN/ZX3VEV I plan to provide Python code as well but I need to learn Python first.
19
102
13
711
145,122
Michael Sachs retweeted
Just gave a guest lecture on Bayesian Causal Inference at Williams College, with slides and R code at doi.org/10.7910/DVN/JOXVQK which is an introduction to our review paper royalsocietypublishing.org/d… @fabri_mealli @FanLiDuke (I never taught any Bayesian Statistics at Berkeley.)
51
1
191
34,590
Michael Sachs retweeted
The New Zealand Herald has an interesting interview with statistical geneticist Bruce Weir after the death of O.J. Simpson. Weir famously made a mistake in his calculations, which defense used to argue that there was bias against Simpson in his methods.🧵 nzherald.co.nz/entertainment…
3
18
1
78
35,741
Have you developed a prediction model to aid clinicians in a treatment decision? Then you may want to read our latest preprint titled *When accurate prediction models yield harmful self-fulfilling prophecies* See: arxiv.org/pdf/2312.01210.pdf A 🧵
1
7
1
23
9,173
if you're writing a paper about a purely predictive model, your paper about it should have to say how those predictions could be useful in real life.
2
2
19
3,923
Michael Sachs retweeted
Derfor skal du spise mere tang: Og sådan høster du selv tang ved de danske strande. #dksund #sundhed #dkvid videnskab.dk/naturvidenskab/…
1
2
7
1,542
Michael Sachs retweeted
In the tidyverse, we work with a lot of people - each other and #rstats community members. We wanted to document how we handle code review, so we've drafted a guide detailing our review principles! We hope you find it useful, and welcome your feedback! tidyverse.github.io/code-rev…
1
34
2
151
38,427
Shiny modules make apps easier to develop, understand, and debug. So why does #rstats treat them as an 'advanced' concept not for beginners? Will discuss in my @appsilon ShinyConf 2023 talk today! It's not too late to join in the fun - like to conf and slides below
1
9
50
8,449
Ever want to add code interactivity to your #quarto or #rmarkdown lecture notes or guides for #rstats? Check out {webr} by @gwstagg and @_lionelhenry for editable and runnable R code. Demo: rd.thecoatlessprofessor.com/… Source in 🧵
3
44
6
138
30,908
📢Why Domain Expertise is important📢 🤨I was looking at a Kaggle Animal Image dataset and saw a picture of a ✨SUSPICIOUSLY✨ fake-looking crab in the dataset. 🦀My student instantly knew it was from a Music Video called "Crab Rave" and confirmed it was a fake crab.
4
5
2
73
13,682
Michael Sachs retweeted
Used R for 20 years, only just discovered base::comment 😂
1
4
412
Michael Sachs retweeted
Paper on "Nested Markov properties for ADMGs" by Thomas Richardson, Jamie Robins, Ilya Shpitser and me was accepted to Annals of Statistics. Read it here! arxiv.org/abs/1701.06686
1
5
1
47
9,519
🧵 In 2019, DeepMind published a paper in Nature describing models to predict AKI in US Veterans, finding AUCs of 92% for RNN & 89% for GBDT, better than prior work. Inspired, we built & tested the generalizability of similar models in & out of VA. Link: rdcu.be/c0OwH
5
43
2
149
Michael Sachs retweeted
Statistikere simulerer VM 100.000 gange - her er deres favoritter til slutrunden: Forskerne giver Danmark 6,5 procents chance for at vinde, mens Brasilien er favoritter. videnskab.dk/kultur-samfund/… #dkvid #dkforsk