Dear CodeSolid Subscriber,
Hello again. US readers, I hope you had an awesome Thanksgiving. Oh, let’s not be exclusive about this: I hope you had an awesome Thanksgiving no matter where you are. Thanksgiving was Thursday.
As many of you know, I’ve been wrestling a bit with the frequency of this newsletter. Once per week is hard to maintain while writing new content, but “hardly ever” isn’t a very good publication schedule either. I think every two weeks fits the bill, so we’ll try that on for a little while.
New On CodeSolid
Recent weeks find me writing a lot about Pandas tools and a bit about alternatives such as Polars.
Large Data Sets in Python: Pandas and the Alternatives
As many of you are probably aware, Pandas is not strictly speaking a big data tool, so for truly massive data sets, something like Spark or Hadoop, or a data warehousing solution like BigQuery, Redshift or the like is generally more appropriate. However, Pandas still works for data sets up to a few gigabytes in size – and this article looks at how to optimize Pandas at this scale as well as another really exciting tool, Polars.
Using SQL With Pandas: PandasSQL, and IPython-SQL with DuckDB
As you may know, Pandas can natively read and write SQL tables as though they were just another data source. Beyond that, there are third-party libraries that let you use SQL to query Pandas DataFrames. One product that supports this brilliantly is DuckDB, which a friend of mine mentioned to me as a general-purpose OLAP tool. The Pandas support was one unexpected bonus of DuckDB. The other fringe benefit was the picture I could use to accompany the article. Let’s face it – few things are cuter than a baby duck!
New Around the Web
How we run our Python tests in hundreds of environments really fast
This article shows how one team sped up their Python test suite, which needed to run hundreds of tests against different frameworks and different Python versions.
Python Asyncio: The Complete Guide
Jason Brownlee of Superfast Python has published a complete guide to programming using the asyncio module.
Pandas DataFrame Visualization Tools
Chris Moffitt of Practical Business Python has written this interesting review of several different tools for viewing Pandas Datasets.
(Just for fun) Vent: I’m tired of the 1001 Libraries of Virtual Environments
If you really want to get a Python discussion board chiming in, just talk about whether all you need is venv and Pip or Conda (or hey what about Poetry?).
I have to admit that I’m also a beneficiary of this confusing state of affairs since my own article on Conda vs. Pip is one of my five most popular articles of all time this month.