@mindpeeker/flowRandomness & order
Flow
Transfer entropy and information dynamics for discrete symbol streams, with χ² and surrogate significance tests.
Transfer entropy asks a directional question about two streams: how many bits does the past of X tell you about the next symbol of Y, beyond what Y’s own past already tells you? Everything below is built from series with a known answer — a planted copy, a hidden driver, a memory at one lag — so you can watch the estimator recover what was put there, and watch it report flow where there is none. Finite samples always give a positive value; read it against a null, never alone.
Loading the demo — every computation runs in your browser.