| | Building a model is one thing. Using its decisions and checking its work is another. In this webinar, Snowplow and Conductrics will build predictive audiences from real time behavioral data, show how to constrain the model, and bring its decisions back into your data for analysis. If you can't make the time, sign up for the recording!
*Message from this week's sponsor, Snowplow |
| | |
| | 🔗 Note: if you don’t see links in this email, it’s also available on the web: dataelixir.com/archive/issue-592.html |
| | |
| | If you use Python for modeling and R for exploring results and making figures, it helps to have both on equal footing. This walkthrough shows how to use both in Positron, with a shared conda environment and a remote setup that keeps the work where the data live.
Mike Talbot |
| | |
| | There are a variety of ways to classify text, from logistic regression to prompting an LLM. This is a great guide to how those approaches work, where a decision model like Jev fits, and the tradeoffs between training your own classifier and using a model that handles different tasks out of the box.
Ahead of AI | Sebastian Raschka, PhD |
| | |
| | This collection of interactive lessons turns the most useful parts of a classic signals and systems textbook into 16 self-contained labs on smoothing, sampling, stability, frequency response, and other ideas that show up in real data work.
regrok | Vipin Gupta |
| | |
| | Repeat the same experiment with the same real effect, and you can get a tiny p-value one time and an unremarkable one the next. This simulation-based walkthrough shows how that happens and why the confidence interval tells you more than the p-value does.
Tidy Ecology |
| | |
| | | | Ledge is an open-source Markdown notebook with code execution built in. Each note gets its own persistent shell and code blocks can run locally or over SSH. It supports Python, SQL, a CLI, and an MCP server for agents to read and update your notes.
ledge.sh |
| | |
| | webrarian is a new R package that turns a folder of R scripts and data into a static website where users can run and edit the code in their browsers. That makes it easy to share an analysis because there’s nothing for users to install and no compute server to maintain.
James Balamuta |
| | |
| | | This is a fun rabbit hole about how people measure autumn, from volunteers watching individual trees to satellites tracking entire forests, and why deciding when the colors are “at peak” is harder than it sounds.
Counting Stuff | Randy Au |
| | |
| | | This week's issue sponsored by: | | |
|
