Posit PBC

Quarto helps Python devs bridge the gap between code and communication.  We've collaborated with Keith Galli to create this 6-part live-coding series exploring how you can create stunning HTML, PDF reports, dashboards, slideshows & even full websites directly from your Python code with Quarto. In case you're not familiar, Quarto is an open-source scientific and technical publishing system that combines the simplicity of Markdown with the power of languages like Python (not to mention R, OJS, and Julia). This makes it incredibly versatile and extensible for creating a wide range of outputs. If you work with Python and need to share data-driven insights, you should seriously consider adopting Quarto in your workflows. This series is a great place to start if you'd like a helping hand getting started. Check out the full list of videos here: https://lnkd.in/dxXMQpuB

How did migrating a codebase to Polars reduce costs by 98%? Jeroen Janssens and Thijs Nieuwdorp detailed their work at Xomnia with client Alliander (a major utilities company in the Netherlands) to process vast network data, where a single network could take over five hours and 500GB of RAM. Moving to Polars resulted in a dramatic improvement, with processing time for a single network reducing to four hours (a 20% reduction despite doubling the samples), RAM usage plummeting by 92% to just 40GB, and the entire low voltage grid calculation costing significantly less than the budget. This is in part thanks to: • Benchmarking: To convince stakeholders, they demonstrated the performance benefits of Polars by benchmarking small pieces of code, showing a reduction from 30 seconds to about one second for a translated pandas snippet. • Polars' lazy API for performance gains: Computations are deferred and optimized compared to the eager API. • Strategic caching: Caching or persisting intermediate results prevents redundant computations of heavy operations within the query plan • Polars' simplified API: Thanks to simplified string handling, consistent representation of missing values, and a lack of index, working with Polars meant a more maintainable codebase. Come for the speed, stay for the API! Watch the recording here: https://lnkd.in/esNTVaMN

🚦 Python data scientists, here's an intriguing analysis project! This week's PydyTuesday explores whether April 20th increases traffic fatalities or if previous research connections were statistical artifacts. You can use Python to investigate this contested correlation and discover how visualization choices can dramatically shape data interpretations. Investigate critical questions such as: ⬢ Can you detect any correlations between fatal car crashes and particular days of the year? ⬢ What are the most dangerous days of the year for fatal car crashes in the United States? ⬢ What other factors might help analyze the data in more detail? You can use the cleaning script to download the full dataset. Your Python Toolkit: Positron → Write your code Quarto → Tell your story Connect Cloud → Share your work 🔢 Get the data: https://lnkd.in/gSy7ae5h 🐍 Learn more about Posit's PydyTuesday project: https://lnkd.in/g_VHKkNf 🎥 Posit PydyTuesday Tutorials: https://lnkd.in/gXCnXKbX Post your submissions to LinkedIn or Bluesky with the hashtags #TidyTuesday and #PydyTuesday for a chance for your work to be featured!

We’re thrilled to announce the release of rsample 1.3.0! rsample makes it easy to create resamples for assessing model performance. It is part of the tidymodels framework, a collection of R packages for modeling and machine learning using tidyverse principles. New updates include: • Flexible grouping for bootstrap intervals: calculate bootstrap confidence intervals of different flavors with a grouping variable. • Tidyverse developer day: Improved documentation, examples, error messages, and more from our amazing community members. Learn more on the tidyverse blog! https://lnkd.in/eXn6VwAN

What if sharing data across your team was as easy as "pinning" it? Join us on April 30th at 11 am ET to see how the {pins} package makes it easier to share datasets and models across teams, especially in Databricks. In this month's Posit Team Demo, Edgar Ruiz will: 📌 Show how to publish assets to Databricks Volumes ⚡ Demonstrate distributed model prediction using Spark + a pinned model 🔗 Cover emerging uses of {pins} in collaborative R workflows Whether you're tired of emailing CSVs or looking to boost your ML pipeline, this one's for you. Add the event to your calendar → https://lnkd.in/eUeNPjwY

Building automated reports in Posit Connect? Our previous post explored dynamic user identity with OAuth for personalized data views. But what if that's not what you need? Our new blog post tackles scenarios where dynamic identity falls short: think scheduled Quarto reports or when Connect users lack matching external accounts. We introduce Service Account OAuth integrations: leveraging a secure, static machine identity for Connect to access data in services like Databricks or Salesforce. Discover how this simplifies workflows, streamlines credential management, and provides a robust solution for consistent data access in non-interactive content in the post! Read here: https://lnkd.in/gevgSAz8

Your weekly Python project is here! Post your submissions to LinkedIn or Bluesky with the hashtags #TidyTuesday and #PydyTuesday for a chance for your work to be featured! Initially featured on TidyTuesday in 2020, this TidyTuesday dataset features the Palmer Penguins! These Antarctic friends return to challenge a new generation of data scientists! Waddle into these exploratory questions: ⬢ The Penguins dataset debuted in 2020 - how have your skills evolved since then? Apply your newest skillset to this dataset to showcase what you’ve learned! ⬢ Can you reproduce (and improve) some iconic visualizations this dataset inspired… without looking at the source code? Google image search “palmer penguins plot” to find some inspiration! ⬢ What clustering patterns emerge when analyzing bill dimensions by island or species? ⬢ Are there enough rows in this data set to predict a penguin’s species accurately? Posit tools you can use: ⬢ Positron → Write your code ⬢ Quarto → Tell your story ⬢ Connect Cloud → Share your work 🐧 Get the data: https://lnkd.in/gcxbHMHh 🙋🏻♀️ Learn more about Posit's PydyTuesday project: https://lnkd.in/g_VHKkNf ▶️ PydyTuesday Tutorials on YouTube: https://lnkd.in/gXCnXKbX

We are thrilled to announce Air, an extremely fast R formatter! A code formatter automates code layout by enforcing predefined style guidelines for whitespace, line breaks, and punctuation. While preserving the code's functionality, it ensures consistent formatting, leading to improved readability, adherence to length constraints, better visual organization of syntax, and reduced merge conflicts in team-based projects. Written in Rust, Air seamlessly integrates with your editor, offering "Format on Save" and "Format Selection" features. Its command-line interface supports formatting individual files, recursive directory formatting, and a --check mode ideal for Git pre-commit hooks or GitHub Actions workflows. Learn more about Air on the tidyverse blog: https://lnkd.in/gNBZdDQB

Stop waiting, start innovating. Want to see your data scientists become agile and independent? Posit Connect Cloud removes the IT barriers, allowing them to deploy and share their work easily. It's like giving them a magic wand that turns data into actionable insights without the IT wait. Posit Connect Cloud lets your team: ☁️ Easily deploy and share data products built with any framework or language. 👷 Eliminate the need for complex infrastructure management. 🔒 Ensure the security and reliability of your data science deployments. Learn more at pos.it/connect_cloud_01

Our roundup of key product updates and new releases, posit::glimpse(), is now out! Learn about some of the amazing tools we've been working on, such as: • LLM-powered tools for R and Python: Querychat lets you embed an SQL-powered LLM in your Shiny app, Ellmer now supports enterprise integration through connectcreds, Chatlas offers a flexible Python interface to a wide range of LLM providers, and Ragnar brings Retrieval-Augmented Generation (RAG) to R. • Python + R, working better together: The reticulate package now supports UV and Gander helps you inspect Python objects from R. We also highlight some recent community events and learning resources, such as: • A three-part workshop walking through building a complex, polished table from scratch using great_tables. • Hadley Wickham's personal retrospective on the origins and evolution of the Tidyverse. We'd love for you to check it out: https://lnkd.in/g9WazbUT