High-Dimensional Changepoint Detection

Efficient implementations of the following multiple changepoint detection algorithms: Efficient Sparsity Adaptive Change-point estimator by Moen, Glad and Tveten (2023) , Informative Sparse Projection for Estimating Changepoints by Wang and Samworth (2017) , and the method of Pilliat et al (2023) .

High-Dimensional Change-point Detection

HDCD contains efficient implementations of several multiple change-point detection algorithms, including Efficient Sparsity Adaptive Change-point estimator (ESAC) and Informative sparse projection for estimating change-points (Inspect).

Installation

You can install the development version of HDCD from GitHub with:

# install.packages("devtools")
devtools::install_github("peraugustmoen/HDCD")

Example

This is a basic example which shows you how to run ESAC:

library(HDCD)
n = 50
p = 50
set.seed(100)
# Generating data
X = matrix(rnorm(n*p), ncol = n, nrow=p)
# Adding a single sparse change-point (at location \eta = 25):
X[1:5, 26:n] = X[1:5, 26:n] +2

# Vanilla ESAC:
res = ESAC(X)
res$changepoints
#> [1] 25

# Manually setting leading constants for \lambda(t) and \gamma(t)
res = ESAC(X,
           threshold_d = 2, threshold_s = 2, #leading constants for \lambda(t)
           threshold_d_test = 2, threshold_s_test = 2 #leading constants for \gamma(t)
)
res$changepoints #estimated change-point locations
#> [1] 25

# Empirical choice of thresholds:
res = ESAC(X, empirical = TRUE, N = 100, tol = 1/100)
res$changepoints
#> [1] 25


# Manual empirical choice of thresholds (equivalent to the above)
thresholds_emp = ESAC_calibrate(n,p, N=100, tol=1/100)
res = ESAC(X, thresholds_test = thresholds_emp[[1]])
res$changepoints
#> [1] 25