loess.fitter: Fit LOESS with K-fold Cross-validation
In arkansas-water-center/TAFA: Trend Analysis with Flow Adjustment

Use LOESS, a local regression smoother, to model the CvQ relationship in the data.

1
2
3

loess.fitter(x, y, span.vals = seq(0.1, 1, by = 0.05),
  folds = control$folds, degree = control$degree, iternum,
  loss = control$loss)

`x`	A vector of the independent variable (i.e., log-Q)
`y`	A vector of the dependent variable (i.e., log-C)
`span.vals`	Sequence of possible values of smoothing parameter (span or "f") to evaluate using cross-validation.
`folds`	The number of folds, or random partitions of data, to use in cross-validation.
`degree`	The degree polynomial to use in LOESS. Defaults to linear (degree = 1)
`iternum`	Current iteration index. Used with loess.opt to make sure the C-V is repeatable for evaluating each span value.
`loss`	The loss function to use. Defaults to mean absolute deviation (MAD); can also use mean squared error (MSE), see details.