Supersmoother vs LOWESS

[ncooder]

You use the stats::supsmu() for smoothing and extracting the trend. Have you tried also LOWESS algorithm and benchmark your solution?

[timginker]

Setting default hyperparameters for LOWESS that work reasonably well in various practical scenarios is challenging. That's why I opted for stats::supsmu() to facilitate automation. However, I agree that adding more options for trend extraction could be beneficial.

[ncooder]

@timginker That is correct, the proper use of LOWESS can be challenging, but still it would be great to have more trend extraction options available in the future.