repr_matrix: Computation of matrix of representations from matrix of time...

In TSrepr: Time Series Representations

Description

The repr_matrix computes matrix of representations from matrix of time series

Usage

repr_matrix(
  x,
  func = NULL,
  args = NULL,
  normalise = FALSE,
  func_norm = norm_z,
  windowing = FALSE,
  win_size = NULL
)

Arguments

x	the matrix, data.frame or data.table of time series, where time series are in rows of the table
func	the function that computes representation
args	the list of additional (or required) parameters of func (function that computes representation)
normalise	normalise (scale) time series before representations computation? (default is FALSE)
func_norm	the normalisation function (default is norm_z)
windowing	perform windowing? (default is FALSE)
win_size	the size of the window

Details

This function computes representation to an every row of a matrix of time series and returns matrix of time series representations. It can be combined with windowing (see repr_windowing) and normalisation of time series.

Value

the numeric matrix of representations of time series

Author(s)

Peter Laurinec, <[e-mail]>

See Also

repr_windowing, repr_list

Examples

# Create random matrix of time series
mat_ts <- matrix(rnorm(100), ncol = 10)
repr_matrix(mat_ts, func = repr_paa,
 args = list(q = 5, func = meanC))

# return normalised representations, and normalise time series by min-max normalisation
repr_matrix(mat_ts, func = repr_paa,
 args = list(q = 2, func = meanC), normalise = TRUE, func_norm = norm_min_max)

# with windowing
repr_matrix(mat_ts, func = repr_feaclip, windowing = TRUE, win_size = 5)

TSrepr documentation built on July 13, 2020, 9:07 a.m.