select links from 2025-11-08
30 Day Map Challenge, filter_out(), making people work for free
Feeling under the weather the past few days but instead of resting I decided that Cities: Skylines was a better remedy. Turns out, I was simply craving the chance to build something that snowballs as you go along. To be honest, this blog is similar - I’m feeling more comfortable with the structure and I’m developing a sense of how each post should look like. I have a soft goal of writing a post in an hour too so the better the sense, the faster the blog post.
It’s been 4 weeks but I strongly recommend writing something for leisure - it’s definitely daunting but writing is an exercise in thinking. If you want to train your muscles, you go to the gym. If you want to train your thinking, well, you’re out of luck. Writing is the gym of the mind.
30 Day Map Challenge
I feel like a bonehead for not sharing this last week on November FIRST. But in case you were not aware, November is the month of challengers such as NanoWrimo or no shave November. This includes the 30 Day Map Challenge - a challenge of producing 30 different maps. Yesterday’s topic was accessibility so I followed Kyle Walker’s excellent tutorial to build an accessibility map during noon and rush hours:
Try making a map yourself! My favourite map bloggers are Kyle Walker and Milos Makes Maps.
Developments in the tidyverse
filter_out()
The tidyverse folks are planning on introducing filter_out() - an inverse of filter(). If you ever needed to exclude rows based on your condition rather than keep them, this one’s for you. Libby’s explainer is lovely:
For a more in-depth discussion, here’s the GitHub issue.
genzplyr
Have your younger colleagues refused to use dplyr? Well, how about genzplyr so they could yeet() instead of filter()!
The docs as well as the pull requests are funny.
Founders Guide: how to get smart people to work for you for free
Here’s a neat blog post by theahura:
If you’re a manager, you’re kind of a tiny company within the larger workplace. This is definitely true for data teams - if you don’t deliver, what’s stopping the business from outsourcing your services? In terms of hiring, what makes an attractive data team? What would make someone want to come work for you for free?
I’ve been thinking about this question this week and so far I’ve come up with the following:
Tools: are your tools modern enough and right for the job? If you do ETL’s using SSIS or Power Automate, good luck finding someone eager to maintain the solutions. Accordingly, if you use LLMs right now, people will be eager to join because that’s an in-demand skill.
Best practices: how “elite” is your team based on DORA metrics? Do you use CI/CD for deployment? Do you maintain internal packages and do you tailor dev tools?
Value preposition: how valued is your team? If it were disbanded today, how much would that affect the overall business? My colleagues LOVE working on problems that other teams deeply care about or that directly drive some top-level metric.
Culture: this one’s tough to pin down but if you create an environment that fosters creativity, teamwork and learning, you’ll have an easier time hiring someone. Developing team culture is not a quick affair and requires A LOT of patience and work. I will probably devote a separate blog post on just this topic alone.
What would have to be true for someone to join your team for no pay?
Other links
Tutorial: UpSet Plots in the Tidyverse by Gleb Ebert
Essay: The Amazon Weekly Business Review (WBR) by Cedric Chin
Book: Crafting Engineering Strategy by Will Larson
Tutorial: TMDL View in Power BI Desktop by FataiSanni
Video/Podcast: Hacking Proteins with AI by Works in Progress