2020#53 Readings of the Week
3 minutes read | 467 words by Ruben BerenguelWe have alternated Friday’s off, so I have extra time to write this post today.
You can also get these as a weekly newsletter by subscribing here.
Python mmap: Improved File I/O With Memory Mapping
Depending on your use cases using mmap
can speed up processing manyfold.
🍿 The Better Boarding Method Airlines Won’t Use
Although I’m a listener of CGP Grey’s Cortex podcast, this is just the 3rd video of him I watch. It’s worth it, it even has a paper as reference. But not like we are going to fly that much anyway…
🔊 Dark Matter
It starts being a bit too predictable, but the last third is pretty entertaining. I found some vibes from Stapledon’s Starmaker (a brilliant and bizarre classic), and Asimov’s The End of Eternity and his short stories What If… and Blank!. Entertaining.
Optimal Peanut Butter and Banana Sandwiches
Mad genius and machine learning. I love this.
Bayes Theorem: A Framework for Critical Thinking
I try to be more Bayesian, but I fail most of the time.
Best Practices Writing Production-Grade PySpark Jobs
As you already know, I mostly use Scala for Spark jobs (sweet, sweet compiler) but for some quick-and-dirty or “if it fails is fine” kind of work I don’t mind using PySpark for the additional speed you can have in development. These suggestions are good to help you package and send.
Property Testing with Complex Inputs
A kind of tutorial for the Hypothesis library for PBT in Python. It will use mypy typing information if you provide it, so if you have used ScalaCheck this will be very familiar.
Hypermodern Python 5: Documentation
I have been reading this series for a while, you can find the previous sections in other Weekly Readings. All of them are good, covering not only the basics and most useful tools but offering good examples to build upon.
Here Dragons Abound: A Meandering Subject
You already should know I love procedural anything and maps in particular. This is a very well explained take on the Meander project, which I have shared before.