Repository navigation
Loess.Function.smooth
commons-math-interpolation / Loess / smooth
function smooth(parms): Float64Array;Defined in: Loess.ts:263
Computes the weighted LOESS linear fit on a sequence of points.
For each point, a local linear regression is computed over the bandwidthFraction part of the points
(with a non-zero weight) that are nearest to it. The points are weighted by the product of their weights
and the tri-cube function of their x distance.
In each robustness iteration, the robustness weights of the points are computed from the residuals
of the previous pass (see SmoothParms.outlierDistanceFactor) and the regression is repeated.
If there are no more than 2 points, a copy of yVals is returned.
The parameters.
Float64Array
The smoothed y values of the points. An element is NaN if the sum of the weights
within its local regression is 0.
Error
If the array lengths do not match, if xVals contains non-finite values or is not monotonically
increasing, if yVals or weights contain non-finite values, if a numeric parameter is invalid,
or if fewer than 2 points have a non-zero weight.