find_motif: Search for Motifs

In tsmp: Time Series with Matrix Profile

Search for Motifs

Description

Search for Motifs

Usage

find_motif(.mp, ...)

## S3 method for class 'MatrixProfile'
find_motif(
  .mp,
  data,
  n_motifs = 3,
  n_neighbors = 10,
  radius = 3,
  exclusion_zone = NULL,
  ...
)

## S3 method for class 'MultiMatrixProfile'
find_motif(
  .mp,
  data,
  n_motifs = 3,
  mode = c("guided", "unconstrained"),
  n_bit = 4,
  exclusion_zone = NULL,
  n_dim = NULL,
  ...
)

## S3 method for class 'PMP'
find_motif(
  .mp,
  data,
  n_motifs = 3,
  n_neighbors = 10,
  radius = 3,
  exclusion_zone = NULL,
  ...
)

Arguments

.mp	a MatrixProfile or MultiMatrixProfile object.
...	further arguments to be passed to class specific function.
data	the data used to build the Matrix Profile, if not embedded.
n_motifs	an int. Number of motifs to find. (Default is 3).
n_neighbors	an int. Number of neighbors to find. (Default is 10).
radius	an int. Set a threshold to exclude matching neighbors with distance > current motif distance * radius. (Default is 3).
exclusion_zone	if a number will be used instead of embedded value. (Default is NULL).
mode	a string. Guided or Unconstrained search. Allow partial match. (Default is guided).
n_bit	an int. Bit size for discretization. Ignored on Guided search. (Default is 4).
n_dim	an int. Number of dimensions to use on Guided search instead of embedded value. (Default is NULL).

Value

For class MatrixProfile, returns the input .mp object with a new name motif. It contains: motif_idx, a list of motif pairs found and motif_neighbor a list with respective motif's neighbors.

For class MultiMatrixProfile, returns the input .mp object with a new name motif. It contains: motif_idx, a vector of motifs found and motif_dim a list the dimensions where the motifs were found

For class PMP, returns the input .mp object with a new name motif. It contains: motif_idx, a list of motif pairs found and motif_neighbor a list with respective motif's neighbors.

Examples

# Single dimension data
w <- 50
data <- mp_gait_data
mp <- tsmp(data, window_size = w, exclusion_zone = 1 / 4, verbose = 0)
mp <- find_motif(mp)

# Multidimension data
w <- mp_toy_data$sub_len
data <- mp_toy_data$data[1:200, ]
mp <- tsmp(data, window_size = w, mode = "mstomp", verbose = 0)
mp <- find_motif(mp)
pan <- tsmp(mp_gait_data, window_size = 20:30, mode = "pmp")
mp <- find_motif(pan)

tsmp documentation built on Aug. 21, 2022, 1:13 a.m.