ISSUE 395  ·   July 12, 2022

Trends

Data teams are getting larger, faster

Data teams are getting bigger, faster but something happens when a team grows past 10 people. You no longer know if the data you use is reliable, the lineage is too large to make sense of and end-users start complaining about data issues every other day. In his latest post, Mikkel Dengsøe explores the issues and ways to handle them as a team scales.
Inside Data | Mikkel Dengsøe

In the News

How likely is it that the audits of Comey and McCabe were a coincidence? A statistical exploration.

It's not often that a news source will teach statistics to expose a story. But without the statistics, you're just left with accusations that could be easy to brand as fake news. How likely is it that the audits were a coincidence? The answer is harder to determine than you might think.
New York Times | Francesca Paris and Josh Katz

Sponsored Link

DoubleCloud: make the most of your data in the cloud

Collect, aggregate and migrate data from multiple sources to build sub-second analytics on fully managed data stacks with proven open-source technologies. Perfect solution for unlocking the potential in your data. Start your trial today

Tutorials, Projects & Opinions

Success Metrics for Product Analytics

Great post on success metrics and how ultimately, they're not a replacement for strategy. They're a way to confirm that the strategy was executed successfully.
The Corner | Daeus Jorento

4 Pandas Anti-Patterns to Avoid and How to Fix Them

pandas users often learn suboptimal coding practices that become their default workflows. This post highlights four common pandas anti-patterns and outlines a set of techniques that you should use instead.
Aidan Cooper

MLOps: Overview, Definition, and Architecture

Great overview of Machine Learning Operations (MLOps) based on a mixed-method research approach, including a literature review, a tool review, and expert interviews. This is a wide-ranging paper that covers MLOps principles, components, roles, architecture, and workflows. There's a lot that's still being figured out in MLOps but for anyone who might think MLOps is simple, check out Figure 4!
arXiv | Dominik Kreuzberger, Niklas Kühl, Sebastian Hirschl

Resources

Python for Data Analysis, 3rd Edition

The third edition of Wes McKinney's Python for Data Analysis has just been released and is free to read online. This is a practical, hands-on guide for manipulating, processing, cleaning, and analyzing datasets with Python. This new edition brings the content up-to-date from 2017.
Wes McKinney

Data in Wonderland

Storytelling can enhance others' understanding of data, especially when combined with analysis and visualization. In this online book, Scott Spencer shows how — covering everything from writing style to visualization & interactives to how to give a good presentation.
Columbia University | Scott Spencer

Data Visualization

Career Portraits in Data Visualization

A lot of people are involved in producing data visualizations. Among analysts, designers, developers, and engineers — who does what, exactly? And what are their backgrounds? This is a great deep dive, based on the Data Visualization Society's "State of the Industry" survey.
Nightingale | Jenn Schilling

Multi-scale model assessment with spatialsample

Nice tutorial that shows how (and why!) to use the new {spatialsample} rstats package to model spatially structured data.
Tidymodels | Mike Mahoney

Outlier

Using GPT-3 to explain how code works

Awesome use-case for GPT-3. Could GPT-3 really explain how sections of code work? Check this out. "It’s shockingly effective."
Simon Willison

How did you like this issue of Data Elixir?

👎    1     2     3     4     5   👍

Was this email forwarded to you?  Sign up here >>