first.step {MissCP}

first.step

Description

Perform the block fused lasso with thresholding to detect candidate break points.

Usage

first.step(
  data_y,
  data_x,
  lambda1,
  lambda2,
  max.iteration = max.iteration,
  tol = tol,
  blocks,
  cv.index,
  fixed_index = NULL,
  nonfixed_index = NULL
)

Arguments

data_y

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

data_x

input data matrix X

lambda1

tuning parmaeter lambda_1 for fused lasso

lambda2

tuning parmaeter lambda_2 for fused lasso

max.iteration

max number of iteration for the fused lasso

tol

tolerance for the fused lasso

blocks

the blocks

cv.index

the index of time points for cross-validation

fixed_index

index for linear regression model with only partial compoenents change.

nonfixed_index

index for linear regression model with only partial compoenents change.

Value

A list object, which contains the followings

jump.l2

estimated jump size in L2 norm

jump.l1

estimated jump size in L1 norm

pts.list

estimated change points in the first step

beta.full

estimated parameters in the first step