stompi_update: Real-time STOMP algorithm

In franzbischoff/tsmp: Time Series with Matrix Profile

Description

Real-time STOMP algorithm

Usage

stompi_update(.mp, new_data, history_size = FALSE)

Arguments

.mp	a TSMP object of class MatrixProfile.
new_data	new data to append to original data.
history_size	an int or FALSE. (Default is FALSE). Keep only this amount of data in the object. The value is for the data, not the matrix profile. Notice that the lmpand lpi will be inconsistent when repeatedly updating limiting the history size and thus will affect the mp and pi.

Value

Returns the input .mp updated with the new information.

Examples

mp <- tsmp(mp_toy_data$data[1:200, 1], window_size = 30, verbose = 0)
mpi <- stompi_update(mp, mp_toy_data$data[201:300, 1])
mp <- tsmp(mp_toy_data$data[1:300, 1], window_size = 30, verbose = 0)
all.equal(mp, mpi, check.attributes = FALSE)

franzbischoff/tsmp documentation built on March 9, 2020, 6:01 a.m.