Contents Up Previous Next

Delta test

Purpose
Description
Macro Synopsis
Modules
Related Functions
References


Purpose

Delta test for detecting dependencies between two time series.

Description

The Delta test is designed to detect dependencies between a group of signals and a reference signal. Furthermore it is suited to give an upper limit for the noise in the reference signal. On the one hand the test is based on the common -definition of continuity, on the other hand on the correlation integral.
An arbitrary number of signals are taken as input signals. The first input signal is considered the reference signal. "Delta test" calculates an dependability index for each of the other signals with respect to the reference signal. These indices are returned in the output signal y. The total dependability and the upper Gauss noise limit of the reference signal are displayed in the message window, together with the dependability indices.
The only parameter is
"Greedy" and "Threshold Accepting" are optimisation algorithms for finding the set of independent variables representing the reference time series with a certain trade-off between computation speed and accuracy. The Greedy-algorithm works by successively building up the vector of independent variables until it is complete and thus the embedding dimension is reached. The Threshold Accepting algorithm performs a local / neighbourhood search based on Simulated Annealing.
Note that since this function operates on a 'pure' time series, the scale and the shift of the given signal do not affect the result.

Macro Synopsis

y = DeltaTest(x,algo);
signal x[];
signal y;
option algo;

Modules

Nonlinear

Related Functions

BDS test, Conditional coupling divergence (PCCD), Mutual cross information, Pointwise transinformation, Post event scan, Synchronicity histogram, Transinformation.

References

Pi/Peterson [25]