Function reference • EWSmethods

Early warning signal assessment

Functions for calculating univariate and multivariate early warning signals.

uniEWS(): Univariate Early Warning Signal Assessment

multiEWS(): Multivariate Early Warning Signal Assessment

perm_rollEWS(): Significance Testing of Rolling Window Early Warning Signals

plot(<EWSmethods>): Plot an EWSmethods object

EWSNet

Functions for interfacing with EWSNet.

ewsnet_init(): EWSNet Initialisation

ewsnet_finetune(): EWSNet Finetune

ewsnet_predict(): EWSNet Predict

ewsnet_reset(): Reset EWSNet Model Weights

conda_clean(): Python Removal

default_weights_path(): Path to Model Weights

S-EWSNet

Functions for interfacing with S-EWSNet.

sewsnet_predict(): S-EWSNet Predict

sewsnet_reset(): Reset S-EWSNet Model

default_sewsnet_path(): Path to S-EWSNet Model

Information imbalance

Functions for estimating Granger causality via information imbalance.

II(): Information Imbalance

tuneII(): Information Imbalance Across Alphas

imbalance_gain(): Information Gain

Other stability metrics

Additional functions estimating alternative measures of system stability/resilience.

FI(): Calculate Fisher Information

uniJI(): Univariate S-map Jacobian index function

uni_smap_jacobian(): Univariate S-map Inferred Jacobian

multiJI(): Multivariate S-map Jacobian index function

multi_smap_jacobian(): Multivariate S-map Inferred Jacobian

mvi(): Multivariate Variance Index function

uniAR(): Univariate Jacobian Index Estimated From an Univariate Autocorrelation Matrix

multiAR(): Multivariate Jacobian Index Estimated From a Multivariate Autocorrelation Matrix

Data preparation

Functions useful for preprocessing data prior to analysis.

detrend_ts(): Detrend Time Series

deseason_ts(): Deseason Seasonal Time Series

embed_ts(): Construct an Embedded Timeseries

Datasets

The exemplary datasets used in EWSmethods tutorials and testing.

simTransComms: Three Simulated Transitioning Communities.

CODrecovery: Three Recovering Cod Populations