Stefano
@stippe87.bsky.social
Former theoretical physicist, now data scientist.
Bayesian, PyMC addicted.
Science outreach enthusiast and dataviz lover.
Skeptik
Mostly in Turin (IT)
https://thestippe.github.io/statistics
Bayesian, PyMC addicted.
Science outreach enthusiast and dataviz lover.
Skeptik
Mostly in Turin (IT)
https://thestippe.github.io/statistics
Pinned
Stefano
@stippe87.bsky.social
· Nov 20
Statistics
All the posts on statistics
thestippe.github.io
Hi, I made public some of my own notes about Bayesian statistics with pymc and
@arviz.bsky.social
on my github.
I hope you fill find them useful.
I hope I will be able to update them soon with some more advanced (so more fun) stuff.
thestippe.github.io/statistics
@arviz.bsky.social
on my github.
I hope you fill find them useful.
I hope I will be able to update them soon with some more advanced (so more fun) stuff.
thestippe.github.io/statistics
Paradossi di linkedin: offerta di lavoro in cui viene richiesta padronanza della lingua italiana menziona "identificate insight azionabili" fra le responsabilita' del ruolo.
November 11, 2025 at 9:01 AM
Paradossi di linkedin: offerta di lavoro in cui viene richiesta padronanza della lingua italiana menziona "identificate insight azionabili" fra le responsabilita' del ruolo.
Reposted by Stefano
Data Science Weekly - Issue 624, by @DataSciNews open.substack.com/pub/datascie...
Data Science Weekly - Issue 624
Curated news, articles and jobs related to Data Science, AI, & Machine Learning
open.substack.com
November 6, 2025 at 11:57 PM
Data Science Weekly - Issue 624, by @DataSciNews open.substack.com/pub/datascie...
Last post on causal inference: DID
Plus: I finally added the "copy to clipboard" button 😁
thestippe.github.io/statistics/d...
Plus: I finally added the "copy to clipboard" button 😁
thestippe.github.io/statistics/d...
Difference in difference
Causal inference from 1850
thestippe.github.io
November 6, 2025 at 9:08 PM
Last post on causal inference: DID
Plus: I finally added the "copy to clipboard" button 😁
thestippe.github.io/statistics/d...
Plus: I finally added the "copy to clipboard" button 😁
thestippe.github.io/statistics/d...
Reposted by Stefano
The Netherlands Food and Consumer Product Authority at the Netherlands Food and Consumer Product Authority is looking for an applied statistician with expertise in Bayesian statistics or causal inference
statmodeling.stat.columbia.edu/2025/11/03/t...
statmodeling.stat.columbia.edu/2025/11/03/t...
The Netherlands Food and Consumer Product Authority at the Netherlands Food and Consumer Product Authority is looking for an applied statistician with expertise in Bayesian statistics or causal infer...
statmodeling.stat.columbia.edu
November 3, 2025 at 9:31 PM
The Netherlands Food and Consumer Product Authority at the Netherlands Food and Consumer Product Authority is looking for an applied statistician with expertise in Bayesian statistics or causal inference
statmodeling.stat.columbia.edu/2025/11/03/t...
statmodeling.stat.columbia.edu/2025/11/03/t...
Reposted by Stefano
This post looks great for someone with solid Bayes skills who wants to work for public good. Requires Dutch language proficiency, but if you speak German and English in my experience you can already understand like 30% of spoken Dutch so could learn it pretty fast!
The Netherlands Food and Consumer Product Authority at the Netherlands Food and Consumer Product Authority is looking for an applied statistician with expertise in Bayesian statistics or causal inference
statmodeling.stat.columbia.edu/2025/11/03/t...
statmodeling.stat.columbia.edu/2025/11/03/t...
The Netherlands Food and Consumer Product Authority at the Netherlands Food and Consumer Product Authority is looking for an applied statistician with expertise in Bayesian statistics or causal infer...
statmodeling.stat.columbia.edu
November 4, 2025 at 7:21 AM
This post looks great for someone with solid Bayes skills who wants to work for public good. Requires Dutch language proficiency, but if you speak German and English in my experience you can already understand like 30% of spoken Dutch so could learn it pretty fast!
Few days ago the head of the DS dept of an IT company told me that he was able to predict fraudulent clients with ML because he had 90% precision in the test set. Seriously?!?
IMHO you can only say this when you know a lot about software and algorithms but nothing about science and human behavior.
IMHO you can only say this when you know a lot about software and algorithms but nothing about science and human behavior.
November 4, 2025 at 8:13 AM
Few days ago the head of the DS dept of an IT company told me that he was able to predict fraudulent clients with ML because he had 90% precision in the test set. Seriously?!?
IMHO you can only say this when you know a lot about software and algorithms but nothing about science and human behavior.
IMHO you can only say this when you know a lot about software and algorithms but nothing about science and human behavior.
On LinkedIn I read someone saying that he was measuring technical debt.
OK, cool. But is it a meaningful measure? Is it stable? Is it robust enough?
If you aren't able to prove this, then what you are selling is just an expensive random number generator.
OK, cool. But is it a meaningful measure? Is it stable? Is it robust enough?
If you aren't able to prove this, then what you are selling is just an expensive random number generator.
October 30, 2025 at 2:35 PM
On LinkedIn I read someone saying that he was measuring technical debt.
OK, cool. But is it a meaningful measure? Is it stable? Is it robust enough?
If you aren't able to prove this, then what you are selling is just an expensive random number generator.
OK, cool. But is it a meaningful measure? Is it stable? Is it robust enough?
If you aren't able to prove this, then what you are selling is just an expensive random number generator.
Reposted by Stefano
We wrote here why we shouldn't even call such unprincipled data dredging or fishing for statistically significant effects "exploratory research". It's time we start taking scientific exploration seriously. Again.
October 28, 2025 at 3:32 PM
We wrote here why we shouldn't even call such unprincipled data dredging or fishing for statistically significant effects "exploratory research". It's time we start taking scientific exploration seriously. Again.
Molte delle grandi compagnie di AI usano pesantemente Python e supportano Trump
Trump, come condizione per donare finanziamenti governativi, richiede che non vengano intraprese azioni per promuovere diversita' ed inclusivita'.
(segue ->)
Trump, come condizione per donare finanziamenti governativi, richiede che non vengano intraprese azioni per promuovere diversita' ed inclusivita'.
(segue ->)
as a condition of funding, we were asked to affirm that we wouldn’t undertake any diversity, equity, and inclusion work, whether or not we used the government funds to do so. The PSF simply couldn’t agree to that statement,
October 28, 2025 at 7:58 AM
Molte delle grandi compagnie di AI usano pesantemente Python e supportano Trump
Trump, come condizione per donare finanziamenti governativi, richiede che non vengano intraprese azioni per promuovere diversita' ed inclusivita'.
(segue ->)
Trump, come condizione per donare finanziamenti governativi, richiede che non vengano intraprese azioni per promuovere diversita' ed inclusivita'.
(segue ->)
Reposted by Stefano
TLDR; The PSF has made the decision to put our community and our shared diversity, equity, and inclusion values ahead of seeking $1.5M in new revenue. Please read and share. pyfound.blogspot.com/2025/10/NSF-...
🧵
🧵
The official home of the Python Programming Language
www.python.org
October 27, 2025 at 2:47 PM
TLDR; The PSF has made the decision to put our community and our shared diversity, equity, and inclusion values ahead of seeking $1.5M in new revenue. Please read and share. pyfound.blogspot.com/2025/10/NSF-...
🧵
🧵
Reposted by Stefano
I'm an #RStats users, but I know some of the R 📦s I use have reticulate + Python under the hood. Thank you @python.org for standing up for everyone in your community. I've made a small donation, hopefully a lot of small $$ can help fund what's needed.
TLDR; The PSF has made the decision to put our community and our shared diversity, equity, and inclusion values ahead of seeking $1.5M in new revenue. Please read and share. pyfound.blogspot.com/2025/10/NSF-...
🧵
🧵
The official home of the Python Programming Language
www.python.org
October 28, 2025 at 2:38 AM
I'm an #RStats users, but I know some of the R 📦s I use have reticulate + Python under the hood. Thank you @python.org for standing up for everyone in your community. I've made a small donation, hopefully a lot of small $$ can help fund what's needed.
A friend of mine told me my latest post was published on the last issue of @datascienceweekly.bsky.social
So cool, thanks!
datascienceweekly.substack.com/p/data-scien...
So cool, thanks!
datascienceweekly.substack.com/p/data-scien...
Data Science Weekly - Issue 622
Curated news, articles and jobs related to Data Science, AI, & Machine Learning
datascienceweekly.substack.com
October 28, 2025 at 7:20 AM
A friend of mine told me my latest post was published on the last issue of @datascienceweekly.bsky.social
So cool, thanks!
datascienceweekly.substack.com/p/data-scien...
So cool, thanks!
datascienceweekly.substack.com/p/data-scien...
Reposted by Stefano
L’OMS ha chiuso 2.000 strutture sanitarie perché i paesi membri hanno ridotto i fondi per gli aiuti internazionali
L’OMS ha chiuso 2.000 strutture sanitarie perché i paesi membri hanno ridotto i fondi per gli aiuti internazionali
ilpost.link
October 24, 2025 at 3:24 PM
L’OMS ha chiuso 2.000 strutture sanitarie perché i paesi membri hanno ridotto i fondi per gli aiuti internazionali
Reposted by Stefano
In a crisis, do you gamble on fast but imperfect data?
Principles for statisticians under pressure: significancemagazine.com/statistics-u...
Principles for statisticians under pressure: significancemagazine.com/statistics-u...
October 24, 2025 at 2:00 PM
In a crisis, do you gamble on fast but imperfect data?
Principles for statisticians under pressure: significancemagazine.com/statistics-u...
Principles for statisticians under pressure: significancemagazine.com/statistics-u...
Reposted by Stefano
Thanks 😄 Actually, @nathanielforde.bsky.social ported this implementation into causalpy causalpy.readthedocs.io/en/stable/no... Check it out :)
Instrumental Variable Modelling (IV) with pymc models — CausalPy 0.5.0 documentation
causalpy.readthedocs.io
October 22, 2025 at 3:10 PM
Thanks 😄 Actually, @nathanielforde.bsky.social ported this implementation into causalpy causalpy.readthedocs.io/en/stable/no... Check it out :)
Instrumental variables with PyMC.
You should also take a look at the brautiful blog post from @juanitorduz.bsky.social where the model is taken, I just adapted it to the well known cigarettes sales.
thestippe.github.io/statistics/i...
You should also take a look at the brautiful blog post from @juanitorduz.bsky.social where the model is taken, I just adapted it to the well known cigarettes sales.
thestippe.github.io/statistics/i...
Instrumental variable regression
Making causal inference without randomization
thestippe.github.io
October 22, 2025 at 8:18 AM
Instrumental variables with PyMC.
You should also take a look at the brautiful blog post from @juanitorduz.bsky.social where the model is taken, I just adapted it to the well known cigarettes sales.
thestippe.github.io/statistics/i...
You should also take a look at the brautiful blog post from @juanitorduz.bsky.social where the model is taken, I just adapted it to the well known cigarettes sales.
thestippe.github.io/statistics/i...
Reposted by Stefano
🧵 Trump, figlio di Dio: la religione come legittimazione del potere e dell’odio | @marcoarvati.bsky.social
www.valigiablu.it/trump-figlio...
www.valigiablu.it/trump-figlio...
October 20, 2025 at 8:51 AM
🧵 Trump, figlio di Dio: la religione come legittimazione del potere e dell’odio | @marcoarvati.bsky.social
www.valigiablu.it/trump-figlio...
www.valigiablu.it/trump-figlio...
Reposted by Stefano
Outrage grows over Trump’s 'disgusting' AI response to 'No Kings'
Outrage grows over Trump’s 'disgusting' AI response to 'No Kings'
Donald Trump shared an AI-generated video of himself dumping feces from a fighter jet onto protestors, following US-wide demonstrations against his administration. This latest - but not uncommon - use of AI from the Trump administration has drawn widespread criticism.
l.euronews.com
October 20, 2025 at 11:21 AM
Outrage grows over Trump’s 'disgusting' AI response to 'No Kings'
Reminder: l'altro ieri Conte diceva che "Essendo presidente, se si fosse dimessa lui l'avrebbe saputo, ma non ne sapeva nulla". Ne concludo che, o è un pessimo presidente, o un disonesto.
Chiara Appendino si è dimessa da vicepresidente del Movimento 5 Stelle
Chiara Appendino si è dimessa da vicepresidente del Movimento 5 Stelle
ilpost.link
October 18, 2025 at 11:56 AM
Reminder: l'altro ieri Conte diceva che "Essendo presidente, se si fosse dimessa lui l'avrebbe saputo, ma non ne sapeva nulla". Ne concludo che, o è un pessimo presidente, o un disonesto.
Reposted by Stefano
‘No Kings’ protest live updates: millions expected to gather across the US for anti-Trump protests
‘No Kings’ protest live updates: millions expected to gather across the US for anti-Trump protests
Organizers expect more than 2,500 demonstrations across country after high turnout at June protests as Robert De Niro urges Americans to ‘stand up and be counted’
Organisers say there are more than 2,500 protests planned across the country. Take a look at where they are expected to take place:
The protests start at different times depending on location. The No Kings website has details for each location. Continue reading...
www.theguardian.com
October 18, 2025 at 11:46 AM
‘No Kings’ protest live updates: millions expected to gather across the US for anti-Trump protests