Matheus Facure Alves

After a year filled with intense writing, I can't believe Causal Inference in Python is finally going to print!! Not only that, but I must say the book looks absolutely beautiful! You can now preorder it on Amazon: amazon.com/Causal-Inferen…

@PDerome2162 Thanks for the kind words!

@MatheusFacure Your book is so freaking good man! Thank you for making such useful information free online! Love from Quebec!

After a year filled with intense writing, I can't believe Causal Inference in Python is finally going to print!! Not only that, but I must say the book looks absolutely beautiful! You can now preorder it on Amazon: amazon.com/Causal-Inferen…

@EpiEllie Sleep. Seems to be one of those topics which is hard to tell the good from the bad science.

There’s a lot of bad public health advice out there, so I’ve started a newsletter to help you sort the information from disinformation. What topics do you want to see covered?

Huge thanks to @MatheusFacure for teaching me the majority of the applied econometrics I know through his excellent free book... matheusfacure.github.io/python-causali…

@PossebomVitor I really like @xuyiqing lectures on panel data for this purpose: youtube.com/playlist?list=…

A friend of mine is looking for an online course on Causal Inference or Policy Evaluation. Optimally, it would be a course at the advanced undergraduate level. #EconTwitter, do you have any recommendations?

@DuquesaDetax Meu segredo é um crianço de 2 anos. 5 am. Nunca falha.

Eu quero que vcs, que acordam super de madrugada, saibam que eu invejo vcs do fundo do meu coração. Vcs e o povo que enche o prato de salada feliz com isso. Pelo amor de Deus, acordar muito cedo é desumano. Café com Deus pai 5am é o 🤬, eu quero tomar café com a Namaria Braga 10

💳💳💳💳 I recently finished a revision of my paper on how credit card lenders set credit limits: 💳💳💳💳

@jonwithno_h Than its easy to transition to the Quini: Take the cumulative effect and multiply by the proportional sample size, so the effect of the top 1% you multiply by 0.01; the effect of the top 2% you multiply by 0.02 and so on.

@jonwithno_h I usually start explaining the cumulative effect: 1) rank order units by what we think predict their sensitivity to the treatment; 2) Compute the effect (in this case, coef of lin. regr) of the top 1%, than top 2% and so on util you get to 100%, where you should have the ATE.

Two gifs for explaining the cumulative gain (quini) and effect curves when the treatment is continuous.

And two more to show what a baseline (stupid model that just outputs random numbers) looks like:

O Tweet da SELIC a 5% rendeu tanta interação boa que realmente não tem nenhuma razão pra não defender SELIC a 5%.

@rphlbruce wtf

Works like a charm. We had loads of fun building a farm and now we're having a blast not knowing wtf we are doing with a bunch o magic spells.

A math department at a major university in the US is now offering a semester long course on “game theory”. Yes, you read that right. Games. The things tiny children play. The dumbing down of America continues…

Oooh boy

😆

Your periodic reminder that economists come in different shapes and sizes. There are academic economists. There are government & central bank economists. There are think tank economists. There are private-sector economists. Each come with useful perspectives & tools.

virgin api consumer vs chad scraper

The only 3 “self-help” books you will ever need