second.step {MissCP}

second.step

Description

Reimputate the missing values and perform the exhaustive search to "thin out" redundant break points.

Usage

second.step(
  data_y,
  data_x,
  max.iteration = max.iteration,
  tol = tol,
  cp.first,
  beta.est,
  blocks,
  data_y_miss
)

Arguments

data_y

input data matrix, with each column representing the time series component

data_x

input data matrix

max.iteration

max number of iteration for the fused lasso

tol

tolerance for the fused lasso

cp.first

the selected break points after the first step

beta.est

the estiamted parameters by block fused lasso

blocks

the blocks

data_y_miss

the data y matrix before the first imputation

Value

A list object, which contains the followings

cp.final

a set of selected break point after the exhaustive search step

beta.hat.list

the estimated coefficient matrix for each segmentation