Delay Differential Analysis

A time series classification framework based on delay differential equations. DDA can be used to extract spatiotemporal dynamics from arbitrary time series.

Streams OpenNeuro ds003029, patient pt01, into DDALAB in your browser. Nothing is uploaded.

u(t) u(t − τ)
Synthetic signal: the x coordinate of a Rössler system, integrated in your browser. Move the delay to unfold the attractor.

One small equation per window

Fitted window by windowLeast squares gives a1, a2, a3 and the fit error for each window of a recording.

A map of the dynamicsOver a recording, the coefficients form a map of its dynamics, window by window and channel by channel.

Five variantsST on single channels; CT, DE and SY on channel pairs; CD for the direction of the interaction between them.

Run it

In the browser

DDALAB opens EDF, BDF, BrainVision, EEGLAB and text recordings where they are, without uploading them.

Open DDALAB

Python and the command line

pip install ddalab runs the same engine for batch work and from Python.

Install

In your own code

The dda-py, dda-rs and DelayDifferentialAnalysis.jl packages.

Packages

Newest papers

All papers