Stabilize the center and scale of a timeseries using robust regression of
DCT bases on the first and second moments.
rob_scale(x, center = TRUE, scale = TRUE, lmrob_method = "MM", rescale = TRUE)
Arguments
| x |
The timeseries to stabilize. |
|---|
| center, scale |
Center and scale? Default: TRUE for both. If
scaling but not centering, the data must already be centered; otherwise,
the results will be invalid. Can also be the number of DCT bases to use for
robust stabilization of center/scale; TRUE will use 4. |
|---|
| lmrob_method |
The lmrob_method argument to robustbase::lmrob. |
|---|
| rescale |
After stabilizing x, re-center and re-scale
to the original mean and variance? Default: TRUE. |
|---|
Value
the timeseries with its center and scale stabilized