Overview

feasts feasts-package

feasts: Feature Extraction and Statistics for Time Series

Graphics

Visualisation is often the first step in understanding the patterns in time series data. The package uses ggplot2 to produce customisable graphics to visualise time series patterns.

autoplot(<tbl_cf>)

Auto- and Cross- Covariance and -Correlation plots

gg_season()

Seasonal plot

gg_subseries()

Seasonal subseries plots

gg_tsdisplay()

Ensemble of time series displays

gg_tsresiduals()

Ensemble of time series residual diagnostic plots

gg_lag()

Lag plots

gg_arma()

Plot characteristic ARMA roots

gg_irf()

Plot impulse response functions

Decompositions

Useful for decomposing a time series into some simpler structural components.

classical_decomposition()

Classical Seasonal Decomposition by Moving Averages

STL()

Multiple seasonal decomposition by Loess

generate(<stl_decomposition>)

Generate block bootstrapped series from an STL decomposition

X_13ARIMA_SEATS()

X-13ARIMA-SEATS Seasonal Adjustment

Autocorrelation analysis

Identify autocorrelations in the data.

ACF() PACF() CCF()

(Partial) Autocorrelation and Cross-Correlation Function Estimation

feat_acf()

Autocorrelation-based features

feat_pacf()

Partial autocorrelation-based features

Unit root tests

Unit root tests for use with fabletools::features().

unitroot_kpss() unitroot_pp()

Unit root tests

unitroot_ndiffs() unitroot_nsdiffs()

Number of differences required for a stationary series

Portmanteau tests

Statistical tests for examining the null hypothesis of independence in a given time series.

ljung_box() box_pierce() portmanteau_tests

Portmanteau tests

Cointegration tests

Statistical tests for investigating cointegration between time series.

cointegration_johansen()

Johansen Procedure for VAR

cointegration_phillips_ouliaris()

Phillips and Ouliaris Cointegration Test

Tiling window features

Computes feature of a time series based on tiled (non-overlapping) windows.

var_tiled_var() var_tiled_mean()

Time series features based on tiled windows

Sliding window features

Computes feature of a time series based on sliding (overlapping) windows.

shift_level_max() shift_var_max() shift_kl_max()

Sliding window features

Other features

Uncategorised features

feat_stl()

STL features

feat_spectral()

Spectral features of a time series

feat_intermittent()

Intermittency features

stat_arch_lm()

ARCH LM Statistic

n_crossing_points()

Number of crossing points

longest_flat_spot()

Longest flat spot length

coef_hurst()

Hurst coefficient

guerrero()

Guerrero's method for Box Cox lambda selection