Greg Brandt

Make performance comparable to R / Fortran

20

On a timeseries of 89345 points with seasonality 8736 (hourly over ~10 yr), it takes 159.25s to compute STL, whereas in R it takes 1.37s. This may be due to...

We should add a test that generates some series with known trend + seasonal and noise / anomalies, then see if we can decompose and get something similar w/in error...