Anytime univariate STAMP algorithm Parallel version

Source: R/stamp_par.R

Computes the best so far Matrix Profile and Profile Index for Univariate Time Series.

stamp.par(..., window.size, exclusion.zone = 1/2, s.size = Inf,
  n.workers = 2, verbose = 2)

Arguments

...

a matrix or a vector. If a second time series is supplied it will be a join matrix profile.
window.size

an int. Size of the sliding window.
exclusion.zone

a numeric. Size of the exclusion zone, based on query size (default is 1/2). See details.
s.size

a numeric. for anytime algorithm, represents the size (in observations) the random calculation will occur (default is Inf).
n.workers

an int. Number of workers for parallel. (Default is 2).
verbose

an int. See details. (Default is 2).

Value

Returns the matrix profile mp and profile index pi. It also returns the left and right matrix profile lmp, rmp and profile index lpi, rpi that may be used to detect Time Series Chains (Yan Zhu 2018).

Details

The Matrix Profile, has the potential to revolutionize time series data mining because of its generality, versatility, simplicity and scalability. In particular it has implications for time series motif discovery, time series joins, shapelet discovery (classification), density estimation, semantic segmentation, visualization, rule discovery, clustering etc. The anytime STAMP computes the Matrix Profile and Profile Index in such manner that it can be stopped before its complete calculation and return the best so far results allowing ultra-fast approximate solutions. verbose changes how much information is printed by this function; 0 means nothing, 1 means text, 2 means text and sound. exclusion.zone is used to avoid trivial matches; if a query data is provided (join similarity), this parameter is ignored.

References

  • Yeh CCM, Zhu Y, Ulanova L, Begum N, Ding Y, Dau HA, et al. Matrix profile I: All pairs similarity joins for time series: A unifying view that includes motifs, discords and shapelets. Proc - IEEE Int Conf Data Mining, ICDM. 2017;1317–22.

  • Zhu Y, Imamura M, Nikovski D, Keogh E. Matrix Profile VII: Time Series Chains: A New Primitive for Time Series Data Mining. Knowl Inf Syst. 2018 Jun 2;1–27.

Website: http://www.cs.ucr.edu/~eamonn/MatrixProfile.html

See also

mstomp(), mstomp.par()

Other Stamp: stamp

Examples

mp <- stamp.par(toy_data$data[1:200,1], window.size = 30, verbose = 0)
# NOT RUN {
ref.data <- toy_data$data[,1]
query.data <- toy_data$data[,2]
# self similarity
mp <- stamp.par(ref.data, window.size = 30, s.size = round(nrows(ref.data) * 0.1))
# join similarity
mp <- stamp.par(ref.data, query.data, window.size = 30, s.size = round(nrows(query.data) * 0.1))
# }