Leonard Stimpfle

PyFixest Sprint

We will be doing a 3-day PyFixest sprint in two weeks - are there any features / regression-based methods widely useful to applied researchers you'd like us to explore/ add?

Yes! Plus git repositories for the underlying code/data free for anyone to fork and play around with. One can dream after all...

What are your thoughts on how an alternative (publication) model should look like?

Similar issues exist for paper and chemicals installations which receive free allocation for heat that may be produced in separate CHP plants.

Great overview, thanks! One comment re Figure 2: The aggregations based on activity types can be misleading because they fail to capture free allocation boundaries. For example, steel is short since 2013 when accounting for transfers of waste gases from blast furnaces to power plants.

"To ascertain the influence of ChatGPT outside the academic context and beyond scripted speech, we analyzed 771,591 podcast episodes spanning diverse topics."

The authors do acknowledge this channel: "despite 56% of YouTube academic talks containing conversational elements, these linguistic shifts stem primarily from direct machine-to-human transmission—such as speakers reading from ChatGPT-co-authored materials"

Along similar lines, @robinlinacre.bsky.social recently shared his take on "respectful" AI use which really resonated with me

It would be very cool if this helped keeping published work up to date at low cost. I thought this was also interesting from a replication/reproducibility perspective more generally

Here's proof that Claude Code can now write an entire empirical polisci paper. Over the weekend, I had Claude Code fully replicate and extend an old paper of mine estimating the effect of univer... Here's proof that Claude Code can now write an entire empirical polisci paper. Over the weekend, I had Claude Code fully replicate and extend an old paper of mine estimating the effect of universal ...

Thoughts on this? www.linkedin.com/posts/andrew...

If you are aware of any other high quality implementations of "modern" estimators in Python, please let us know!

Errors and Inconsistencies in European Emission Data Factories that suddenly have emissions a thousand times lower than before, unexplained data gaps, and inconsistent CO₂ emissions from different data sources are some of the issues one finds when havin...

Was there a massive leak of a dangerous greenhouse gas in Iceland🇮🇸 in 2011-or was it a data📈 reporting error?
Emission databases show that the Norðurál aluminium factory🏭 released 60 tons of SF₆ in 2011 — but no such emissions in any other year. 🔌💡 industrydecarbonization.com/news/errors-...
🧵

Great to finally see Ruff integrated!

Thank you for publishing this! Quick feedback: Clicking on markers in the map of the "Company operations" tab doesn't seem to do anything (I've tried both Safari and Chrome)

The 2021 spike seems to be caused by low generation in the first three months (chart for the Rovinari power plant). Is something wrong with the reported time interval (ResolutionCode)? Any experts on the ENTSO-E generation data who know what might be going on? #EnergySky

All data is publicly available: Industrial Emissions Portal (industry.eea.europa.eu); ENTSO-E (transparency.entsoe.eu); linking between databases (hdl.handle.net/1814/94107). 8/8

Where does this leave us? Oltenia's power plants report suspiciously low emission factors per unit of coal burned. This can be driven by inaccurate reporting of emissions or energy inputs. Emissions per unit of electricity generated appear to be more in line with averages across Europe. 7/

This second chart presents emission factors per unit of energy output. Compared to tCO2 per TJ, Oltenia's power plants have less extreme emission factors in terms of tCO2 per MWh. For example, Isalnita's emission factor is close to or above the average emission factor in most years. 6/

Although electricity production is not reported in the Industrial Emissions Portal, it can be obtained from the ENTSO-E transparency platform (transparency.entsoe.eu) and linked back to the reporting in the Industrial Emissions Portal. 5/

Oltenia claims that this is due to low-quality lignite combusted in its plants. @occrp.org rightly points out that low-quality lignite should result in higher emission factors "per unit of energy produced". However, the article only shows emissions per unit of energy input and not energy output. 4/

The chart presents the mean and 95% confidence interval of the 74 power plants reporting lignite energy inputs since 2016. Although there is a similar trend on average, Oltenia's power plants appear to have some of the lowest emission factors across Europe. 3/

The analysis compares trends in country-level average emission factors in tCO2 per TJ of lignite. Can we observe similar differences at the plant-level? For this, I dug into the reporting of lignite-fired power plants in the Industrial Emissions Portal provided by the @eea.europa.eu

This is a great analysis on a very important topic.

But are the emission factors driven by the numerator (emissions) or denominator (energy inputs)?

A short investigation (cc @hanno.hboeck.de)

These are calculated using generation data from the ENTSO-E transparency platform. With @eui-eu.bsky.social, we have recently published a linking between the Industrial Emissions Portal, the ENTSO-E and the EU ETS Union Registry hdl.handle.net/1814/94107

Emission factors in tCO2 per MWh (top panel) and tCO2 per TJ (bottom panel)

I just checked if the emission factors expressed as tCO2 per MWh show a similar pattern (albeit noisier).

Does the reporting of energy content per unit of coal the Industrial Emissions Portal account for the specific coal used in these plants? In other words, is the trend driven by emissions or energy input reporting?

It would be interesting to see if these papers take a design-based view on inference. From @cdechaisemartin.bsky.social's difference-in-differences book: "We do not adopt the design-based perspective, as it seems unrealistic in many natural experiments." papers.ssrn.com/abstract=448...

Why is it needlessly difficult to access UNFCCC Emission Data? Due to international climate agreements, countries report annual data about their greenhouse gas emissions to the UNFCCC. However, downloading and using that data is a lot more difficult than it shoul...

The UNFCCC makes it needlessly difficult to download emission data📊📈

The UNFCCC🇺🇳 is most famously known for being the organizer of the climate conferences (COPs). However, they also play an important role in collecting and managing emission data. 🔌💡
industrydecarbonization.com/news/why-is-...
🧵

[new draft post] Real purchasing power over time is not economic welfare over time https://drafts.interfluidity.com/2025/11/12/real-purchasing-power-over-time-is-not-economic-welfare-over-time/index.html

Screenshot of blog post at https://dirk.eddelbuettel.com/blog/2025/11/11#duckdb-mlpack_0.0.4 describing release 0.0.4 of the duckdb community extension mlpack.

duckdb-mlpack 0.0.4: Added random forest and logistic regression
Bringing mlpack machine learning to duckdb SQL
dirk.eddelbuettel.com/blog/2025/11...

Adds random forest and logistic regression, reworks interface, adds new state variable for verbosity, expands documentation and adds unit tests