Seemingly smooth distributions can look very different with more bins; always worth checking when using any histogram tool (real data: movie budgets)
In pandas, explode() will turn single rows with a column containing a list of values into multiple rows with a single value (the gif explains it better).
The great thing about standards is that there are so many to choose from :-) nevertheless, glad something like Python Data API Standards is happening buff.ly/3azhe13 most end users will never see this kind of meta-behind-the-scenes work.
KDE plots plus scatter plots on the same small multiple chart? Why not, with @matplotlib #seaborn @pandas_dev and @numpy_team. Find it in chapter 17 of the Biological Data Exploration book pythonforbiologists.com/biol…
A fascinating approach to machine learning: wrap everything in a single function call which takes a natural language description of the problem. Anyone tried this? buff.ly/32Xvr6h
My first attempt at a screencast video: a deep dive into making Manhattan plots with Python, pandas et. al. youtube.com/watch?v=Ej4FkgLO…
Lately I've mostly been using Jupyter widgets to prototype simple GUIs, but this is an interesting alternative buff.ly/31UJLfl
Overview and code for building a bullet chart (I hadn't heard of it before either) buff.ly/2gGedUH
This looks like the ultimate library for terminal colours/effects buff.ly/2pKgKAs
Register by Friday for Python datavis in Berlin. Learn how to make this PCA plot and a zillion others buff.ly/2ln4v89
Cool idea: colourize tracebacks to make it easier to pick out the important lines buff.ly/2cP2BNp
Another charting library with a few interesting twists: Altair buff.ly/2cbzfZF
