Describe the feature or idea you want to propose

thanks to @Cyril-Meyer we have a nice collection of smoothing algorithms. It would be nice to have a simple version of LOESS:

"LOESS (Locally Estimated Scatterplot Smoothing) is a non-parametric regression technique used to produce a smooth curve through noisy data. It was proposed by Cleveland (1979, 1988), and is often employed in exploratory data analysis, time series smoothing, and data visualization because of its flexibility and simplicity."

Describe your proposed solution

The general definition is for x and exogenous variables, but I think we only need the univariate version such as used in STL.

Describe alternatives you've considered, if relevant

No response

Additional context

No response