fusion_estimates: fusion_estimates
In joon3216/funpark: Functions written by Junkyu Park

Description Usage Arguments Value Author(s) See Also Examples

Find the fusion estimates of given y. Start iteration with theta and the tuning parameter lambda.

1	fusion_estimates(y, theta, lambda, max_iter = 100, eps = 1e-5)

`y`	A vector of numbers.
`theta`	A vector of numbers that has the same length as `y`. If missing, then it is replaced with `y`.
`lambda`	A positive number; the tuning parameter.
`max_iter`	A natural number; the maximum iteration. `100` by default.
`eps`	A positive number; `1e-05` by default.

A list:

theta: Resulting fusion estimates.
phi: Differences in theta.
lambda: A specified lambda when running the function.
iter: The number of iterations.
costs: Collection of costs at each iteration.

Junkyu Park

Dealing with a non-separable penalty

set.seed(1024)
n <- 1000
t <- 1:n
f <- function(t) {t / 250 - .5}
g <- function(t) {-(.25 / 449) * t + 250 / 449}
true_theta <- c(rep(0, 249), f(250:500), rep(.75, 50), g(551:1000))
y <- true_theta + rnorm(1000, 0, 0.1)
start <- rep(.5, n)
result <- fusion_estimates(y, theta = Start, lambda = 1, max_iter = 1000)