trendfilter_weights: Solve trendfiltering problem with weights.

In jean997/jadeTF:

Description

This function minimizes

1/2 || w(y - θ) ||^2 + λ_||Dθ||_1

Where D is a trendfiltering pendalty matrix. This function borrows code heavily from the genlasso package and also returns a genlasso object.

Usage

trendfilter_weights(y, pos, wts = NULL, ord = 0, approx = FALSE,
  maxsteps = 2000, minlam = 0, rtol = 1e-07, btol = 1e-07,
  eps = 1e-04, verbose = FALSE)

Arguments

y	Vector of data.
pos	Vector of positions.
wts	Vector of weights
ord	Order of polynomial to fit.
...	See trendfilter for other options.

Value

Returns a genlasso object with additional elements weights with the weights provided.

jean997/jadeTF documentation built on May 18, 2019, 11:44 p.m.