This is certainly because I'm more proficient in Python, but the point is that without a proper controlled experiment, I think we're bound to discuss personal anecdotes that do not necessarily generalize well to other people.

"I would consider the R code to be slightly easier to read (notice how many quotes and brackets the Python code needs) ". Disagree, I find R slightly harder to read, as it's more difficult to distinguish between variables and methods.

What I'd like to see one day is a proper controlled experiment on the effectiveness of using one language vs. the other on a variety of tasks, comparing beginners in each language, and comparing proficient programmers in each language. It would require to think very carefully about the metric used.

"This also means performance considerations are secondary". Disagree with that, when you have to use simulations, performance can be an important issue when you have tight deadlines, possibly making both pure R and pure Python inadequate.

This is not mentioned on the page, but beware of RColorBrewer, some palettes are not colorblind-friendly. However one can identify which ones are, with the command display.brewer.all(colorblindFriendly=TRUE)

In social sciences, there's also the issue of stakeholders' cooperation. Some may want to control the study design, so the results won't reflect badly on them. If their cooperation is necessary for the study to be conducted at all, the issue is not limited to funding or to knowledge of stats.

Joking aside, some researchers might be reluctant to share coauthorship. It reminds me of this bsky.app/profile/phde...

free buses for everyone: not wild, but might be partly disputable if the available financial resources make you choosing between that, or having more buses more frequently. Not necessarily the priority if your bus system is very shitty. But I guess this is not what they're talking about in NYC.

in fact the first version of the graph (without texture) renders better for people with monochromacy. Still, I'd adapt the order of the legend, it might be a bit surprising otherwise (we'd expect "don't know" to be at the top of bars).

As a possible improvement, I'd match the order of colors in the graph to their order in the legend. The color of "don't know" and "don't trust very much" are a bit similar for people with monochromacy, so that might be a slightly tricky to immediately interpret the graph, though the texture helps.

So I wouldn't recommend this website to anyone.

besides H2N2, possibly the 7th cholera pandemic was affecting French colonies in South East Asia at that time. It might explain partly the drop even if far from all of it, but I don't know if there is data about that.

(Perhaps the confusion was that you thought the graph I shared was for the world? It is not, it's for France)

I don't see the relationship between the Great Leap Forward and the drop of life expectancy in France in 1961-63 that we're talking about here (but are we?). Anyway it was to point out that the "80 years" time frame is not correct for France.

There's a drop between 1961 and 1963 that is similar or worse depending on the age group (e.g. 71 years of life expectancy in 1961 to 70.3 in 1963, i.e. a drop of 0.7; 2019-2020 is a drop of 0.5 at birth). Look at ourworldindata.org/grapher/life...

Ce qui me surprend, c'est qu'ils n'aient pas ajouté "et de l'intelligence artificielle" au passage dans le titre.

Chez moi, la même chose s'affiche sur chrome ou sur firefox. Ils ont peut-être modifié la page mais il y a un temps de latence selon où on se trouve, ou alors une histoire de cache.

je suis sur ordinateur, pas sur un téléphone. Bizarre.

Rien ici non plus.