-
Notifications
You must be signed in to change notification settings - Fork 7
Loess
Christian d'Heureuse edited this page Sep 25, 2026
·
1 revision
commons-math-interpolation / Loess
LOESS/LOWESS local regression.
This module implements the LOESS/LOWESS algorithm for local regression (see Local regression). For each point, a weighted linear least-squares fit is computed over the neighboring points. Optional robustness iterations reduce the influence of outliers.
smooth computes the smoothed y values of the points. createLoessInterpolator additionally connects the smoothed points with an interpolation method and returns a function.
| Interface | Description |
|---|---|
| LoessInterpolatorDiagInfo | Diagnostics info returned by createLoessInterpolator. |
| LoessInterpolatorParms | Parameters for createLoessInterpolator. |
| SmoothDiagInfo | Diagnostics info returned by smooth. |
| SmoothParms | Parameters for smooth. |
| Function | Description |
|---|---|
| calculateLocalLinearRegression | Calculates the weighted least-squares linear fit at position x over the points iLeft ... iRight. |
| createLoessInterpolator | Returns a function that interpolates the LOESS-smoothed values of a dataset. |
| smooth | Computes the weighted LOESS linear fit on a sequence of points. |