README.md

In nhart985/NonParamLines: Nonparametric Local Linear Smoothing

NonParamLines

The NonParamLines package provides implementations of two popular nonparametric regression estimators: LOESS (Cleveland, 1979) and Friedman's running lines estimator (Friedman, 1984). The two main functions in this package are loess_fit() and supsmu_fit(), which generate response predictions given the values of a continuous predictor variable. See help pages and vignettes for details and examples.

The R implementations in the NonParamLines package are slower than existing implementations in the stats package, which use Fortran and C++. This is expected since each algorithm require a loop, but all of the code within the loop is vectorized and efficient. See source code for details.

nhart985/NonParamLines documentation built on Dec. 22, 2021, 2:09 a.m.