Bootstrapping with LOWESS
Sample the original data with replacement and fit LOWESS curves accordingly.
boot.lowess(x, y = NULL, f = 2/3, iter = 3, line.col = "#FF000033", ...)
We keep on resampling the data and finally we will see several bootstrapped LOWESS curves, which may give us a rough idea about a “confidence interval” of the LOWESS fit.
Yihui Xie
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.