pacoramon

How do you benchmark causal inference methods when you need both realistic data and pre-specified causal ground truth? We introduce CausalMix: a generative sandbox that balances realism with causal controllability. Paper: arxiv.org/abs/2603.03587 Code: github.com/zhangqiecho/ca…

Creating reports by hand every time your data changes can slow you down significantly. Each update often means adjusting tables, rerunning analyses, and rebuilding visualizations manually. This takes time and makes it easy for inconsistencies to slip in. Quarto in R provides a much more efficient approach. It allows you to write your analysis, text, and code in one place. When you render the document, all results, tables, and plots are generated automatically based on your current data. Everything stays consistent without extra effort. Once your report is set up, updating it is simple. Just refresh the data and render the file again. The example below shows how code and output are directly connected, making the entire workflow transparent and reproducible. I recently added a new module to the Statistics Globe Hub that walks through this process step by step. It includes a full video, practical examples, and exercises to help you build your own automated reports in R. Not part of the Statistics Globe Hub yet? The Hub is an ongoing program with a new module every week on statistics, data science, AI, R, and Python. Join now to access the Quarto module and all other modules released this month. More information about the Statistics Globe Hub: statisticsglobe.com/hub #rstats #datascience #statistics #quarto #reporting #statisticsglobehub

📢 Hiring: Project Lead, Data Scientist Join the @OECD Economics Department’s new data-focused unit and lead innovative work at the intersection of AI, data science and policy. 👉 Interested? Apply: brnw.ch/21x16P7 📅 Deadline: 13 April 2026 (midnight CEST) #AIJobs

📢 Reminder! Submit your abstracts for the Observational Health Data Sciences and Informatics (OHDSI) Conference by 1 May 2026. OHDSI promotes open science and collaboration in health data research through adoption of the OMOP Common Data Model. Whether you would like to present a poster, software demo or lightning talk, we welcome abstracts from across the OMOP community. 📆 17 - 18 September 2026 📍 Nottingham Submit 👉 hubs.li/Q047m98j0 Abstract template 👉 hubs.li/Q047lFJ00

🟣¿Sabes que desde ‘Mi Carpeta Ciudadana' puedes acceder a diversos servicios de la Seguridad Social? ✏️ Averigua el estado de tu solicitud ✏️ Gestiona la renovación de la Tarjeta Sanitaria Europea ✏️ Consulta tu Número de Afiliación 📌 Más información f.mtr.cool/fhhqcoqtxl

🚀 Busquem Analista de Dades Júnior amb impacte social Vols treballar amb dades obertes i contribuir a projectes generant coneixement i transformació social? 📍Barcelona |💼 Indefinit 👉Presenta la teva candidatura forms.gle/dLqELnijtq71xx… #OpenData #ImpacteSocial #DataWomen

Gracias a todos los que pudieron venir (presencial y online). Agradecer de nuevo @emilopezcano y al @DSLAB_URJC por todas las facilidades. El material de la presentación, ya está disponible: madrid.r-es.org/102-martes-24-… @RConsortium - @UNED - @DSLAB_URJC - @R_Hisp

‼️🇪🇸 A database from a Spain-based travel and geolocation software company is allegedly being sold on a cybercrime forum. ▪️ Industry: GeoLocation Software / Travel ▪️ User Regions: U.S., Italy, Norway, Spain, Ireland ▪️ Price: $325 The database contains multiple tables: ▪️ Users Table (170K+ rows): UID, name, pass, local domain, email, theme, signature, signature format, created, access, login, status, timezone, language, picture, init, data, changed, UUID ▪️ Twitter Account Table (310K+ rows): Twitter UID, host, screen name, OAuth tokens, description, location, followers count, friends count, statuses count, favourites count, profile URL, profile image, background/sidebar/border colors, verified status, created at, mentions, last refresh, is global UID ▪️ Geo Address Table (350K+ rows): Entity type, bundle, deleted, entity ID, revision ID, language, delta, country, administrative area, sub-administrative area, locality, dependent locality, postal code, thoroughfare, premise, sub-premise, organisation name, name line, first name, last name, geo address data

‼️🇪🇸 The Spanish electricity company FENIE Energía has allegedly been completely hacked, with a sample of the database released on a popular cybercrime forum after the company reportedly failed to respond to emails. ▪️ Total Size: 430GB ▪️ Total Rows: 1,700,000+ ▪️ Includes: Complete database plus files (ID card images, contracts, cards, etc.) The CSV structure reveals an extensive set of fields including: ▪️ Customer identity data (tipo_id, CIF/NIF, name, surnames) ▪️ Business group and corporate details ▪️ Property administrator and finca data ▪️ Contact info (phone numbers, email, commercial details) ▪️ Fiscal data (CP, province, locality, street, fiscal number, escalera, piso, puerta) ▪️ Billing/invoicing details (invoice address, province, locality, street, floor, door, CP) ▪️ Banking info (IBAN, BIC/SWIFT, IBAN España, bank account, payment method) ▪️ Electronic invoicing and email invoicing fields ▪️ Administrative and public entity data (CUPS, razón social, oficina contable, órgano gestor, unidad tramitadora) ▪️ Dates (fecha_boe, fecha_alta, fecha_baja) The poster notes this is not the complete database and that additional data, columns, and files are withheld.

Did you know? There are 5 options for medical doctors to apply to live and work in Canada permanently: •Express Entry •Provincial Nominee Program •Atlantic Immigration Program •Rural Community Immigration Pilot •Francophone Community Immigration Pilot Explore these programs and how to apply today: bit.ly/49xGh3u

🚀Hoy, Europa y España marcan un hito en la navegación por satélite con el lanzamiento de #CELESTE desde Nueva Zelanda.   🇪🇸 Desde @cienciagob, junto a la @aee_gob, lideramos la participación española y reforzamos nuestro papel en las misiones espaciales europeas.

Según informa el CEO de Solugen, 200 contenedores de helio líquido están atrapados en el Golfo Pérsico sin refrigeración activa, con un margen de máximo 48 días antes de que el gas se pierda irreversiblemente. (...) Dado que el helio ultrapuro es esencial para la litografía EUV y no tiene sustituto, se viene escasez de este gas fundamental para nuestro modelo industrial.

Check out the Statistics Globe Hub: statisticsglobe.com/hub The Statistics Globe Hub is an ongoing learning program that helps you stay up to date with statistics, data science, AI, and programming using R and Python. #quarto #rstats #datascience #reproducibility #reporting #statisticsglobehub

A First Course in Causal Inference: arxiv.org/abs/2305.18793 [490-page PDF download] + Also see the book "Causal Inference in Statistics: A Primer" at amzn.to/3Mrm2wO by @yudapearl #Probability #Mathematics #DataScience #ML #MachineLearning #DataScientist #DataAnalysis

Google Student Researcher 2026 application closes in 2 days!!!!! All students should apply for this, its an amazing opportunity. Application Link : google.com/about/careers/…

Learn how to choose the right data visualizations for different datasets and communication goals, and apply best practices to create clear, impactful visuals. Enroll now! Self-paced. wrld.bg/4Jkb50YtMz0

🚨 $220K. 3 months. No PhD. No gatekeeping. Perplexity's Research Residency is hunting brilliant minds from ANY field: physics, cogsci, quant finance, math, philosophy + more. Bring your unique brain. We'll bring compute, mentorship & mission. Rolling apps. SF/Palo Alto. Visa support. Apply before your genius friend does 👇 perplexity.ai/hub/ai-researc… #AI #BuildWithPerplexity #ResearchResidency @perplexity_ai

Mediation and Moderation Analysis using R: A Complete Guide to Mediation, Moderation, and Causal Analysis Using R and Real-World Data by Alexandru Cernat is the featured course on Leanpub! This course introduces mediation and moderation analysis using R, covering interaction effects, path analysis, SEM, and causal mediation. Gain the skills to separate mediating vs moderating variables, and state-of-the-art methods such as Structural Equation Modelling and causal mediation. #R

There's a lot of discussion about technical debt introduced by "AI slop" code. I'm not convinced by this. In 2015, I wrote the tigris R package. Since then, it's been downloaded over 1 million times from the Posit CRAN mirror, putting it within the top 5 percent of R packages. Not bad! I would not say that the code is brilliant. I'm not a programmer by training. I wanted a tool that automatically downloaded shapefiles from the Census Bureau. It solved my problem. Mostly a brute-forced solution. I wasn't worried about best practices or test coverage, I just wanted it to work. I've had no shortage of ideas; the code was always the bottleneck. Either the time to write it, or my own abilities. With AI, that's no longer the case. I can push my ideas beyond my technical limits because AI is writing my code. I don't know Rust, but I want its high-performance capabilities in my work. I'm now writing R packages with Rust backends - with the help of Claude Code and Codex - that bring my ideas to life in ways well beyond what I could have done on my own. My latest package {freestiler}, a vector tiling engine written in Rust, is a great example. It's powering all of the cool visualizations I've been posting over the past month. Is the code "AI slop?" It's way better than what I would have produced on my own, and it works because I have tested and verified the results. And others are now using and benefiting from the tool - which would not have existed without LLMs! AI-generated code that ships and solves real problems beats code never written and ideas never shared.