September 6, 2022 at 8:35 PM

Big news, seaborn 0.12 has been released.

The biggest change is the object API, inspired by the grammar of graphics.

But I can’t possibly be the only one bothered by one of their examples?

Especially the y0 and y1 feels super implicit and group=“Country”, col=None feels like coming out of an ancient book written on knowledge long forgotten.

Am I ruined by my years using matplotlib during my PhD?

so.Plot(healthexp, ×=“Year” ) .pair(y=[“Spending_USD”, “Life_Expectancy”]) . facet(col=“Country”, order=country order) •scale(x=so. Continuous(). tick(at=[1985, 2020])) . label(x="" y0= “Spending (USD)”, y1=“Life expectancy”) .add(so.Line(color="#ceced3" ), group=“Country”, col=None) .add(so.Line(linewidth=3), color=“Country”, legend=False)

https://michaelwaskom.medium.com/announcing-the-release-of-seaborn-0-12-f26266ddbd8f