Title: Time Series Feature Extraction
Version: 1.1.1
Description: Methods for extracting various features from time series data. The features provided are those from Hyndman, Wang and Laptev (2013) <doi:10.1109/ICDMW.2015.104>, Kang, Hyndman and Smith-Miles (2017) <doi:10.1016/j.ijforecast.2016.09.004> and from Fulcher, Little and Jones (2013) <doi:10.1098/rsif.2013.0048>. Features include spectral entropy, autocorrelations, measures of the strength of seasonality and trend, and so on. Users can also define their own feature functions.
Depends: R (≥ 3.6.0)
Imports: fracdiff, forecast (≥ 8.3), purrr, RcppRoll (≥ 0.2.2), stats, tibble, tseries, urca, future, furrr
Suggests: testthat, knitr, rmarkdown, ggplot2, tidyr, dplyr, Mcomp, GGally
License: GPL-3
ByteCompile: true
URL: https://pkg.robjhyndman.com/tsfeatures/, https://github.com/robjhyndman/tsfeatures
BugReports: https://github.com/robjhyndman/tsfeatures/issues
RoxygenNote: 7.2.3
VignetteBuilder: knitr
Encoding: UTF-8
NeedsCompilation: no
Packaged: 2023-08-28 13:23:56 UTC; robjhyndman
Author: Rob Hyndman ORCID iD [aut, cre], Yanfei Kang ORCID iD [aut], Pablo Montero-Manso [aut], Mitchell O'Hara-Wild ORCID iD [aut], Thiyanga Talagala ORCID iD [aut], Earo Wang ORCID iD [aut], Yangzhuoran Yang [aut], Souhaib Ben Taieb [ctb], Cao Hanqing [ctb], D K Lake [ctb], Nikolay Laptev [ctb], J R Moorman [ctb], Bohan Zhang [ctb]
Maintainer: Rob Hyndman <Rob.Hyndman@monash.edu>
Repository: CRAN
Date/Publication: 2023-08-28 14:00:02 UTC

tsfeatures: Time Series Feature Extraction

Description

Methods for extracting various features from time series data. The features provided are those from Hyndman, Wang and Laptev (2013) doi:10.1109/ICDMW.2015.104, Kang, Hyndman and Smith-Miles (2017) doi:10.1016/j.ijforecast.2016.09.004 and from Fulcher, Little and Jones (2013) doi:10.1098/rsif.2013.0048. Features include spectral entropy, autocorrelations, measures of the strength of seasonality and trend, and so on. Users can also define their own feature functions.

Author(s)

Maintainer: Rob Hyndman Rob.Hyndman@monash.edu (ORCID)

Authors:

Other contributors:

See Also

Useful links:

Autocorrelation at lag 9. Included for completion and consistency.

Description

Autocorrelation at lag 9. Included for completion and consistency.

Usage

ac_9(y, acfv = stats::acf(y, 9, plot = FALSE, na.action = na.pass))

Arguments

y

the input time series

acfv

vector of autocorrelation, if exist, used to avoid repeated computation.

Value

autocorrelation at lag 9

Author(s)

Yangzhuoran Yang

References

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

Autocorrelation-based features

Description

Computes various measures based on autocorrelation coefficients of the original series, first-differenced series and second-differenced series

Usage

acf_features(x)

Arguments

x

a univariate time series

Value

A vector of 6 values: first autocorrelation coefficient and sum of squared of first ten autocorrelation coefficients of original series, first-differenced series, and twice-differenced series. For seasonal data, the autocorrelation coefficient at the first seasonal lag is also returned.

Author(s)

Thiyanga Talagala

ARCH LM Statistic

Description

Computes a statistic based on the Lagrange Multiplier (LM) test of Engle (1982) for autoregressive conditional heteroscedasticity (ARCH). The statistic returned is the R^2 value of an autoregressive model of order lags applied to x^2.

Usage

arch_stat(x, lags = 12, demean = TRUE)

Arguments

x

a univariate time series

lags

Number of lags to use in the test

demean

Should data have mean removed before test applied?

Value

A numeric value.

Author(s)

Yanfei Kang

Convert mts object to list of time series

Description

An mts object contains a multivariate time series in a matrix, with time on rows. This is converted into a list of univariate time series.

Usage

## S3 method for class 'mts'
as.list(x, ...)

Arguments

x

multivariate time series of class mts.

...

other arguments are ignored.

Value

A list of ts objects.

Author(s)

Rob J Hyndman

The autocorrelation feature set from software package hctsa

Description

Calculate the features that grouped as autocorrelation set, which have been used in CompEngine database, using method introduced in package hctsa.

Usage

autocorr_features(x)

Arguments

x

the input time series

Details

Features in this set are embed2_incircle_1, embed2_incircle_2, ac_9, firstmin_ac, trev_num, motiftwo_entro3, and walker_propcross.

Value

a vector with autocorrelation features

Author(s)

Yangzhuoran Yang

References

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

See Also

embed2_incircle

ac_9

firstmin_ac

trev_num

motiftwo_entro3

walker_propcross

Converts an input vector into a binarized version from software package hctsa

Description

Converts an input vector into a binarized version from software package hctsa

Usage

binarize_mean(y)

Arguments

y

the input time series

Value

Time-series values above its mean are given 1, and those below the mean are 0.

Author(s)

Yangzhuoran Yang

References

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

CompEngine feature set

Description

Calculate the features that have been used in CompEngine database, using method introduced in package hctsa.

Usage

compengine(x)

Arguments

x

the input time series

Details

The features involved can be grouped as autocorrelation, prediction, stationarity, distribution, and scaling.

Value

a vector with CompEngine features

Author(s)

Yangzhuoran Yang

References

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

See Also

autocorr_features

pred_features

station_features

dist_features

scal_features

Number of crossing points

Description

Computes the number of times a time series crosses the median.

Usage

crossing_points(x)

Arguments

x

a univariate time series

Value

A numeric value.

Author(s)

Earo Wang and Rob J Hyndman

The distribution feature set from software package hctsa

Description

Calculate the features that grouped as distribution set, which have been used in CompEngine database, using method introduced in package hctsa.

Usage

dist_features(x)

Arguments

x

the input time series

Details

Features in this set are histogram_mode_10 and outlierinclude_mdrmd.

Value

a vector with distribution features

Author(s)

Yangzhuoran Yang

References

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

See Also

histogram_mode

outlierinclude_mdrmd

Points inside a given circular boundary in a 2-d embedding space from software package hctsa

Description

The time lag is set to the first zero crossing of the autocorrelation function.

Usage

embed2_incircle(
  y,
  boundary = NULL,
  acfv = stats::acf(y, length(y) - 1, plot = FALSE, na.action = na.pass)
)

Arguments

y

the input time series

boundary

the given circular boundary, setting to 1 or 2 in CompEngine. Default to 1.

acfv

vector of autocorrelation, if exist, used to avoid repeated computation.

Value

the proportion of points inside a given circular boundary

Author(s)

Yangzhuoran Yang

References

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

Spectral entropy of a time series

Description

Computes spectral entropy from a univariate normalized spectral density, estimated using an AR model.

Usage

entropy(x)

Arguments

x

a univariate time series

Details

The spectral entropy equals the Shannon entropy of the spectral density f_x(\lambda) of a stationary process x_t:

H_s(x_t) = - \int_{-\pi}^{\pi} f_x(\lambda) \log f_x(\lambda) d \lambda,

where the density is normalized such that \int_{-\pi}^{\pi} f_x(\lambda) d \lambda = 1. An estimate of f(\lambda) can be obtained using spec.ar with the burg method.

Value

A non-negative real value for the spectral entropy H_s(x_t).

Author(s)

Rob J Hyndman

References

Jerry D. Gibson and Jaewoo Jung (2006). “The Interpretation of Spectral Entropy Based Upon Rate Distortion Functions”. IEEE International Symposium on Information Theory, pp. 277-281.

Goerg, G. M. (2013). “Forecastable Component Analysis”. Proceedings of the 30th International Conference on Machine Learning (PMLR) 28 (2): 64-72, 2013. Available at https://proceedings.mlr.press/v28/goerg13.html.

See Also

spec.ar

Examples

entropy(rnorm(1000))
entropy(lynx)
entropy(sin(1:20))

Time of first minimum in the autocorrelation function from software package hctsa

Description

Time of first minimum in the autocorrelation function from software package hctsa

Usage

firstmin_ac(
  x,
  acfv = stats::acf(x, lag.max = N - 1, plot = FALSE, na.action = na.pass)
)

Arguments

x

the input time series

acfv

vector of autocorrelation, if exist, used to avoid repeated computation.

Value

The lag of the first minimum

Author(s)

Yangzhuoran Yang

References

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

Examples

firstmin_ac(WWWusage)

The first zero crossing of the autocorrelation function from software package hctsa

Description

Search up to a maximum of the length of the time series

Usage

firstzero_ac(y, acfv = stats::acf(y, N - 1, plot = FALSE, na.action = na.pass))

Arguments

y

the input time series

acfv

vector of autocorrelation, if exist, used to avoid repeated computation.

Value

The first zero crossing of the autocorrelation function

Author(s)

Yangzhuoran Yang

References

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

Longest flat spot

Description

"Flat spots” are computed by dividing the sample space of a time series into ten equal-sized intervals, and computing the maximum run length within any single interval.

Usage

flat_spots(x)

Arguments

x

a univariate time series

Value

A numeric value.

Author(s)

Earo Wang and Rob J Hyndman

Implements fluctuation analysis from software package hctsa

Description

Fits a polynomial of order 1 and then returns the range. The order of fluctuations is 2, corresponding to root mean square fluctuations.

Usage

fluctanal_prop_r1(x)

Arguments

x

the input time series (or any vector)

Author(s)

Yangzhuoran Yang

References

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

Heterogeneity coefficients

Description

Computes various measures of heterogeneity of a time series. First the series is pre-whitened using an AR model to give a new series y. We fit a GARCH(1,1) model to y and obtain the residuals, e. Then the four measures of heterogeneity are: (1) the sum of squares of the first 12 autocorrelations of y^2; (2) the sum of squares of the first 12 autocorrelations of e^2; (3) the R^2 value of an AR model applied to y^2; (4) the R^2 value of an AR model applied to e^2. The statistics obtained from y^2 are the ARCH effects, while those from e^2 are the GARCH effects.

Usage

heterogeneity(x)

Arguments

x

a univariate time series

Value

A vector of numeric values.

Author(s)

Yanfei Kang and Rob J Hyndman

Mode of a data vector from software package hctsa

Description

Measures the mode of the data vector using histograms with a given number of bins as suggestion. The value calculated is different from hctsa and CompEngine as the histogram edges are calculated differently.

Usage

histogram_mode(y, numBins = 10)

Arguments

y

the input data vector

numBins

the number of bins to use in the histogram.

Value

the mode

Author(s)

Yangzhuoran Yang

References

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

Parameter estimates of Holt's linear trend method

Description

Estimate the smoothing parameter for the level-alpha and the smoothing parameter for the trend-beta. hw_parameters considers additive seasonal trend: ets(A,A,A) model.

Usage

holt_parameters(x)

hw_parameters(x)

Arguments

x

a univariate time series

Value

holt_parameters produces a vector of 2 values: alpha, beta.

hw_parameters produces a vector of 3 values: alpha, beta and gamma.

Author(s)

Thiyanga Talagala, Pablo Montero-Manso

Hurst coefficient

Description

Computes the Hurst coefficient indicating the level of fractional differencing of a time series.

Usage

hurst(x)

Arguments

x

a univariate time series. If missing values are present, the largest contiguous portion of the time series is used.

Value

A numeric value.

Author(s)

Rob J Hyndman

The first zero crossing of the autocorrelation function of the residuals from Simple local time-series forecasting from software package hctsa

Description

Simple predictors using the past trainLength values of the time series to predict its next value.

Usage

localsimple_taures(y, forecastMeth = c("mean", "lfit"), trainLength = NULL)

Arguments

y

the input time series

forecastMeth

the forecasting method, default to mean. mean: local mean prediction using the past trainLength time-series values. lfit: local linear prediction using the past trainLength time-series values.

trainLength

the number of time-series values to use to forecast the next value. Default to 1 when using method mean and 3 when using method lfit.

Value

The first zero crossing of the autocorrelation function of the residuals

Time series features based on tiled windows

Description

Computes feature of a time series based on tiled (non-overlapping) windows. Means or variances are produced for all tiled windows. Then stability is the variance of the means, while lumpiness is the variance of the variances.

Usage

lumpiness(x, width = ifelse(frequency(x) > 1, frequency(x), 10))

stability(x, width = ifelse(frequency(x) > 1, frequency(x), 10))

Arguments

x

a univariate time series

width

size of sliding window

Value

A numeric vector of length 2 containing a measure of lumpiness and a measure of stability.

Author(s)

Earo Wang and Rob J Hyndman

Time series features based on sliding windows

Description

Computes feature of a time series based on sliding (overlapping) windows. max_level_shift finds the largest mean shift between two consecutive windows. max_var_shift finds the largest var shift between two consecutive windows. max_kl_shift finds the largest shift in Kulback-Leibler divergence between two consecutive windows.

Usage

max_level_shift(x, width = ifelse(frequency(x) > 1, frequency(x), 10))

max_var_shift(x, width = ifelse(frequency(x) > 1, frequency(x), 10))

max_kl_shift(x, width = ifelse(frequency(x) > 1, frequency(x), 10))

Arguments

x

a univariate time series

width

size of sliding window

Details

Computes the largest level shift and largest variance shift in sliding mean calculations

Value

A vector of 2 values: the size of the shift, and the time index of the shift.

Author(s)

Earo Wang and Rob J Hyndman

Local motifs in a binary symbolization of the time series from software package hctsa

Description

Coarse-graining is performed. Time-series values above its mean are given 1, and those below the mean are 0.

Usage

motiftwo_entro3(y)

Arguments

y

the input time series

Value

Entropy of words in the binary alphabet of length 3.

Author(s)

Yangzhuoran Yang

References

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

Examples

motiftwo_entro3(WWWusage)

Nonlinearity coefficient

Description

Computes a nonlinearity statistic based on Lee, White & Granger's nonlinearity test of a time series. The statistic is 10X^2/T where X^2 is the Chi-squared statistic from Lee, White and Granger, and T is the length of the time series. This takes large values when the series is nonlinear, and values around 0 when the series is linear.

Usage

nonlinearity(x)

Arguments

x

a univariate time series

Value

A numeric value.

Author(s)

Yanfei Kang and Rob J Hyndman

References

Lee, T. H., White, H., & Granger, C. W. (1993). Testing for neglected nonlinearity in time series models: A comparison of neural network methods and alternative tests. Journal of Econometrics, 56(3), 269-290.

Teräsvirta, T., Lin, C.-F., & Granger, C. W. J. (1993). Power of the neural network linearity test. Journal of Time Series Analysis, 14(2), 209–220.

Examples

nonlinearity(lynx)

How median depend on distributional outliers from software package hctsa

Description

Measures median as more and more outliers are included in the calculation according to a specified rule, of outliers being furthest from the mean.

Usage

outlierinclude_mdrmd(y, zscored = TRUE)

Arguments

y

the input time series (ideally z-scored)

zscored

Should y be z-scored before computing the statistic. Default: TRUE

Details

The threshold for including time-series data points in the analysis increases from zero to the maximum deviation, in increments of 0.01*sigma (by default), where sigma is the standard deviation of the time series.

At each threshold, proportion of time series points included and median are calculated, and outputs from the algorithm measure how these statistical quantities change as more extreme points are included in the calculation.

Outliers are defined as furthest from the mean.

Value

median of the median of range indices

Author(s)

Yangzhuoran Yang

References

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

Partial autocorrelation-based features

Description

Computes various measures based on partial autocorrelation coefficients of the original series, first-differenced series and second-differenced series

Usage

pacf_features(x)

Arguments

x

a univariate time series

Value

A vector of 3 values: Sum of squared of first 5 partial autocorrelation coefficients of the original series, first differenced series and twice-differenced series. For seasonal data, the partial autocorrelation coefficient at the first seasonal lag is also returned.

Author(s)

Thiyanga Talagala

The prediction feature set from software package hctsa

Description

Calculate the features that grouped as prediction set, which have been used in CompEngine database, using method introduced in package hctsa.

Usage

pred_features(x)

Arguments

x

the input time series

Details

Features in this set are localsimple_mean1, localsimple_lfitac, and sampen_first.

Value

a vector with prediction features

Author(s)

Yangzhuoran Yang

References

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

See Also

localsimple_taures

sampen_first

Second Sample Entropy of a time series from software package hctsa

Description

Modified from the Ben Fulcher's EN_SampEn which uses code from PhysioNet. The publicly-available PhysioNet Matlab code, sampenc (renamed here to RN_sampenc) is available from: http://www.physionet.org/physiotools/sampen/matlab/1.1/sampenc.m

Usage

sampen_first(y)

Arguments

y

the input time series

Details

Embedding dimension is set to 5. The threshold is set to 0.3.

Author(s)

Yangzhuoran Yang

References

cf. "Physiological time-series analysis using approximate entropy and sample entropy", J. S. Richman and J. R. Moorman, Am. J. Physiol. Heart Circ. Physiol., 278(6) H2039 (2000)

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

Second Sample Entropy from software package hctsa

Description

Modified from the Ben Fulcher version of original code sampenc.m from http://physionet.org/physiotools/sampen/ http://www.physionet.org/physiotools/sampen/matlab/1.1/sampenc.m Code by DK Lake (dlake@virginia.edu), JR Moorman and Cao Hanqing.

Usage

sampenc(y, M = 6, r = 0.3)

Arguments

y

the input time series

M

embedding dimension

r

threshold

Author(s)

Yangzhuoran Yang

References

cf. "Physiological time-series analysis using approximate entropy and sample entropy", J. S. Richman and J. R. Moorman, Am. J. Physiol. Heart Circ. Physiol., 278(6) H2039 (2000)

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

The scaling feature set from software package hctsa

Description

Calculate the features that grouped as scaling set, which have been used in CompEngine database, using method introduced in package hctsa.

Usage

scal_features(x)

Arguments

x

the input time series

Details

Feature in this set is fluctanal_prop_r1.

Value

a vector with scaling features

Author(s)

Yangzhuoran Yang

References

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

See Also

fluctanal_prop_r1

Bootstrap-based stationarity measure from software package hctsa

Description

100 time-series segments of length l are selected at random from the time series and the mean of the first zero-crossings of the autocorrelation function in each segment is calculated.

Usage

spreadrandomlocal_meantaul(y, l = 50)

Arguments

y

the input time series

l

the length of local time-series segments to analyse as a positive integer. Can also be a specified character string: "ac2": twice the first zero-crossing of the autocorrelation function

Value

mean of the first zero-crossings of the autocorrelation function

Author(s)

Yangzhuoran Yang

References

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

The stationarity feature set from software package hctsa

Description

Calculate the features that grouped as stationarity set, which have been used in CompEngine database, using method introduced in package hctsa.

Usage

station_features(x)

Arguments

x

the input time series

Details

Features in this set are std1st_der, spreadrandomlocal_meantaul_50, and spreadrandomlocal_meantaul_ac2.

Value

a vector with stationarity features

Author(s)

Yangzhuoran Yang

References

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

See Also

std1st_der

spreadrandomlocal_meantaul

Standard deviation of the first derivative of the time series from software package hctsa

Description

Modified from SY_StdNthDer in hctsa. Based on an idea by Vladimir Vassilevsky.

Usage

std1st_der(y)

Arguments

y

the input time series. Missing values will be removed.

Value

Standard deviation of the first derivative of the time series.

Author(s)

Yangzhuoran Yang

References

cf. http://www.mathworks.de/matlabcentral/newsreader/view_thread/136539

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

Strength of trend and seasonality of a time series

Description

Computes various measures of trend and seasonality of a time series based on an STL decomposition. The number of seasonal periods, and the length of the seasonal periods are returned. Also, the strength of seasonality corresponding to each period is estimated. The mstl function is used to do the decomposition.

Usage

stl_features(x, ...)

Arguments

x

a univariate time series.

...

Other arguments are passed to mstl.

Value

A vector of numeric values.

Author(s)

Rob J Hyndman

Normalized nonlinear autocorrelation, the numerator of the trev function of a time series from software package hctsa

Description

Calculates the numerator of the trev function, a normalized nonlinear autocorrelation, The time lag is set to 1.

Usage

trev_num(y)

Arguments

y

the input time series

Value

the numerator of the trev function of a time series

Author(s)

Yangzhuoran Yang

References

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

Examples

trev_num(WWWusage)

Time series feature matrix

Description

tsfeatures computes a matrix of time series features from a list of time series

Usage

tsfeatures(
  tslist,
  features = c("frequency", "stl_features", "entropy", "acf_features"),
  scale = TRUE,
  trim = FALSE,
  trim_amount = 0.1,
  parallel = FALSE,
  multiprocess = future::multisession,
  na.action = na.pass,
  ...
)

Arguments

tslist

a list of univariate time series, each of class ts or a numeric vector. Alternatively, an object of class mts may be used.

features

a vector of function names which return numeric vectors of features. All features returned by these functions must be named if they return more than one feature. Existing functions from installed packages may be used, but the package must be loaded first. Functions must return a result for all time series, even if it is just NA.

scale

if TRUE, time series are scaled to mean 0 and sd 1 before features are computed.

trim

if TRUE, time series are trimmed by trim_amount before features are computed. Values larger than trim_amount in absolute value are set to NA.

trim_amount

Default level of trimming if trim==TRUE.

parallel

If TRUE, multiple cores (or multiple sessions) will be used. This only speeds things up when there are a large number of time series.

multiprocess

The function from the future package to use for parallel processing. Either multisession or multicore. The latter is preferred for Linux and MacOS.

na.action

A function to handle missing values. Use na.interp to estimate missing values.

...

Other arguments get passed to the feature functions.

Value

A feature matrix (in the form of a tibble) with each row corresponding to one time series from tslist, and each column being a feature.

Author(s)

Rob J Hyndman

Examples

mylist <- list(sunspot.year, WWWusage, AirPassengers, USAccDeaths)
tsfeatures(mylist)

Unit Root Test Statistics

Description

unitroot_kpss computes the statistic for the Kwiatkowski et al. unit root test using the default settings for the ur.kpss function. unitroot_pp computes the statistic for the Phillips-Perron unit root test using the default settings for the ur.pp function.

Usage

unitroot_kpss(x, ...)

unitroot_pp(x, ...)

Arguments

x

a univariate time series.

...

Other arguments are passed to the ur.kpss or ur.kpss functions.

Value

A numeric value

Author(s)

Pablo Montero-Manso

Simulates a hypothetical walker moving through the time domain from software package hctsa

Description

The hypothetical particle (or 'walker') moves in response to values of the time series at each point. The walker narrows the gap between its value and that of the time series by 10%.

Usage

walker_propcross(y)

Arguments

y

the input time series

Value

fraction of time series length that walker crosses time series

Author(s)

Yangzhuoran Yang

References

B.D. Fulcher and N.S. Jones. hctsa: A computational framework for automated time-series phenotyping using massive feature extraction. Cell Systems 5, 527 (2017).

B.D. Fulcher, M.A. Little, N.S. Jones Highly comparative time-series analysis: the empirical structure of time series and their methods. J. Roy. Soc. Interface 10, 83 (2013).

Yahoo server metrics

Description

Yahoo server metrics

Usage

yahoo_data(...)

Arguments

...

Additional arguments passed to download.file

Downloads and returns aggregated and anonymized datasets from Yahoo representing server metrics of Yahoo services.

Value

A matrix of time series with 1437 rows of hourly data, and 1748 columns representing different servers.

Author(s)

Rob Hyndman, Earo Wang, Nikolay Laptev, Mitchell O'Hara-Wild

References

Hyndman, R.J., Wang, E., Laptev, N. (2015) Large-scale unusual time series detection. In: Proceedings of the IEEE International Conference on Data Mining. Atlantic City, NJ, USA. 14–17 November 2015. https://robjhyndman.com/publications/icdm2015/

Examples

yahoo <- yahoo_data()
plot(yahoo[,1:10])
plot(yahoo[,1:44], plot.type='single', col=1:44)

Proportion of zeros

Description

Computes proportion of zeros in a time series

Usage

zero_proportion(x, tol = 1e-08)

Arguments

x

a univariate time series

tol

tolerance level. Absolute values below this are considered zeros.

Value

A numeric value.

Author(s)

Thiyanga Talagala