The goal of changepoint.cov is to provide methods for detecting covariance or subspace changepoints in multivariate time series.
You can install the development version from GitHub with:
# install.packages("devtools")
devtools::install_github("grundy95/changepoint.cov")These basic examples show how to use the 2 main functions: cptCov for
detecting changes in covariance and cptSubspace for detecting changes
in subspace.
For detecting covariance changepoints in high-dimensional independent
time series, we recommend using method='Ratio'.
set.seed(1)
data <- wishartDataGeneration(n=200,p=30,tau=100)$data
ansRatio <- cptCov(X=data, method='Ratio')
summary(ansRatio)
#> Created using changepoint.cov version 0.5.0
#> Method : Ratio
#> Multiple changepoint method : AMOC
#> Minimum segment length : 60
#> Changepoints : 96
plot(ansRatio)
For detecting covariance changes in low-dimensional, potentially
dependent time series we recommend using method='CUSUM'.
set.seed(1)
data <- wishartDataGeneration(n=200,p=3,tau=c(50,150))$data
ansCUSUM <- cptCov(X=data, method='CUSUM', numCpts='BinSeg', threshold='Manual', thresholdValue=7)
show(ansCUSUM)
#> Class, cptCovariance : Covariance Changepoint x
#> : S4 class containing 15 slots with names
#> subspaceDim nperm LRCov covEst subspaceEst version date method dataset cpts msl numCpts threshold thresholdValue cptsSig
#>
#> Created on : Tue Mar 30 15:08:42 2021
#>
#> Summary(.) :
#> ---------------
#> Created using changepoint.cov version 0.5.0
#> Method : CUSUM
#> Multiple changepoint method : BINSEG
#> Minimum segment length : 6
#> Changepoints : 59 151
cptsSig(ansCUSUM)
#> cpts T thresholdValue significant
#> 2 59 13.255543 7 TRUE
#> 3 151 7.618705 7 TRUE
#> 4 45 3.834405 7 FALSE
covEst(ansCUSUM)
#> [[1]]
#> PC1 PC2 PC3
#> PC1 0.2411427 0.2113139 -0.1651886
#> PC2 0.2113139 1.3903415 -0.4141982
#> PC3 -0.1651886 -0.4141982 0.2836532
#>
#> [[2]]
#> PC1 PC2 PC3
#> PC1 0.74584049 0.1139156 0.04684264
#> PC2 0.11391557 0.8629125 0.37255546
#> PC3 0.04684264 0.3725555 1.13315173
#>
#> [[3]]
#> PC1 PC2 PC3
#> PC1 0.41821814 -0.03955294 -0.03716283
#> PC2 -0.03955294 0.46688293 0.01584048
#> PC3 -0.03716283 0.01584048 0.65631946For detecting subspace changepoints in high-dimensional data using the
cptSubspace function.
set.seed(1)
data <- subspaceDataGeneration(n=100,p=20,subspaceDim=5,tau=50,changeSize=0.5*sqrt(5))$data
ansSubspace <- cptSubspace(X=data, subspaceDim=5, nperm=100)
summary(ansSubspace)
#> Created using changepoint.cov version 0.5.0
#> Method : Subspace
#> Multiple changepoint method : AMOC
#> Minimum segment length : 20
#> Changepoints : 50
subspaceEst(ansSubspace)
#> [[1]]
#> [,1] [,2] [,3] [,4] [,5]
#> [1,] -0.306634945 -0.38061692 0.03005095 0.209841195 -0.21479599
#> [2,] -0.023295229 0.04601747 0.11817483 0.114644462 -0.07736184
#> [3,] 0.174299174 -0.25667090 0.03613504 -0.171354019 -0.26148231
#> [4,] 0.133216989 -0.12438758 -0.35646986 0.147388756 0.51801321
#> [5,] -0.109341525 0.30684585 -0.01431329 0.044041076 -0.27728919
#> [6,] -0.171235562 -0.07505962 -0.01755111 -0.188008598 -0.01271848
#> [7,] 0.304641803 0.36868876 0.01806936 -0.119612599 -0.04146972
#> [8,] 0.096282782 -0.33368612 -0.45711018 0.037820635 0.07682975
#> [9,] -0.038054730 0.05239461 -0.33548262 -0.001355701 -0.04867948
#> [10,] -0.000965908 -0.38624978 0.02810438 0.166202222 -0.29047239
#> [11,] 0.096625759 0.17251767 0.04881956 0.515146417 -0.17969347
#> [12,] -0.147989968 0.26260145 -0.31655962 -0.040649005 -0.28602655
#> [13,] 0.088885150 -0.08412251 0.01491905 -0.107928755 -0.38868098
#> [14,] -0.248334386 0.04343216 0.24013350 -0.527154576 0.12459279
#> [15,] 0.501940051 0.17939694 -0.29909892 -0.134633464 -0.14811028
#> [16,] 0.520126848 -0.25514769 0.22029173 -0.022657370 -0.03455092
#> [17,] -0.022536327 0.01966256 0.04634592 -0.082834253 0.27935349
#> [18,] -0.279784360 0.14631540 -0.27888880 0.212943674 0.01567108
#> [19,] 0.098898494 0.07234943 0.35944776 0.382147802 0.23890839
#> [20,] 0.025497953 0.20165072 0.16133608 0.194001310 -0.02368223
#>
#> [[2]]
#> [,1] [,2] [,3] [,4] [,5]
#> [1,] 0.23912953 0.088054704 0.08715689 -0.33534499 0.244377967
#> [2,] 0.07756817 0.137447707 0.33138435 -0.04647981 -0.148354192
#> [3,] -0.04106168 0.297907213 -0.07919225 -0.50331402 -0.026211305
#> [4,] 0.15661800 -0.574608455 -0.26878505 0.05044587 0.081807861
#> [5,] -0.44921030 0.001585932 0.30644056 0.07864855 -0.103495912
#> [6,] -0.02567239 -0.309374588 0.17300124 -0.26140297 0.092574265
#> [7,] -0.18746445 0.085666476 -0.14026622 0.17889212 -0.450893221
#> [8,] 0.12464379 -0.210618904 -0.35755113 -0.21960345 -0.282729694
#> [9,] -0.26730980 -0.269036324 -0.11504467 -0.18415632 -0.101433459
#> [10,] 0.17319975 0.022771378 -0.11925605 -0.29763431 -0.205272244
#> [11,] 0.36429767 0.032102668 0.28596499 -0.11940202 -0.277630358
#> [12,] -0.32699784 0.041236572 -0.03228107 -0.31858449 -0.260024912
#> [13,] 0.11029588 0.057495681 0.02477713 -0.12357811 -0.161367972
#> [14,] -0.32035827 0.172551575 -0.23212165 -0.04820899 0.536474447
#> [15,] -0.26416783 -0.081507671 -0.12557837 0.18305050 -0.289118290
#> [16,] 0.18617261 0.246826832 -0.56018838 0.11173160 -0.046916255
#> [17,] 0.01169026 -0.112157618 -0.04248947 0.01490675 0.068938976
#> [18,] -0.04547149 -0.446837625 0.11093726 -0.17842696 0.074941772
#> [19,] 0.12257790 -0.131266954 0.15267950 0.28554157 -0.035231399
#> [20,] 0.28265671 0.035575076 0.03073275 0.24359038 -0.002572745