Wasserstein

Distribution shape, transport cost, self-similarity
distributionaldim distribution space6 metrics

What It Measures

How the distribution of values differs from uniform, and how stable that distribution is across the signal.

Bins the data into a 32-bin histogram and treats it as a probability distribution. Then asks three questions: how far is this distribution from uniform (optimal transport cost)? How concentrated is it (peak height)? And does the first half of the signal look like the second half (self-similarity)?

Metrics

concentration

The peak bin height times the number of bins. 1.0 means uniform (De Bruijn scores exactly 1.0 — its construction guarantees every byte pattern appears equally). Above 1.0 means the distribution has a spike. Collatz gap lengths (31.8), Rainfall (31.5), and Forest fire (29.4) are the most concentrated signals in the atlas — their heavy-tailed distributions pile most of their mass into the lowest bin.

dist_from_uniform

Earth mover's distance from the uniform distribution: the minimum amount of "dirt" you'd need to move to make the histogram flat. Collatz gap lengths (0.48) and Rainfall (0.48) are farthest from uniform. Neural net pruned weights (0.46) are close behind — pruning creates a spike at zero. De Bruijn scores near 0 (already uniform).

entropy

Shannon entropy of the 32-bin histogram. De Bruijn, circle map quasiperiodic, and phyllotaxis all score 5.0 (near the maximum of 5 bits — flat distribution). Collatz gap lengths scores 0.05 (almost all mass in one bin). This is the classical measure of distributional spread, here computed on the Wasserstein embedding.

self_similarity

One minus the earth mover's distance between the first-half and second-half histograms. 1.0 means the distribution is perfectly stable over time (logistic period-4, logistic period-2, De Bruijn). Hilbert walk scores 0.60 (its deterministic sweep creates different distributions in the first and second halves). This catches nonstationarity that entropy and concentration miss: a signal can have high entropy overall but low self_similarity if its distribution drifts.

transport_variability

Coefficient of variation of windowed earth mover's distances between consecutive segments. Exponential Chirp (0.21) scores highest — its frequency sweep creates rapidly changing local distributions. Sunspot (0.13) and Pulse-Width Mod (0.13) also score high. Constants and periodic orbits score 0.0 (identical windows). This measures how much the optimal transport cost fluctuates over time — a windowed nonstationarity detector complementing self_similarity's global split. Evolved via ShinkaEvolve.

recurrence_distance

Average earth mover's distance between non-adjacent windows that are within a recurrence threshold. PID Controller (0.11) and Exponential Chirp (0.10) score highest — their recurring distributional states differ in fine detail. Constants score 0.0. This measures how similar the signal's distribution is when it "returns" to a previously visited distributional state. Evolved via ShinkaEvolve.

Atlas Rankings

concentration
SourceDomainValue
Mian-Chowlanumber_theory31.8488
Aubry-André Criticalquantum31.7109
Rainfall (ORD Hourly)climate31.4687
···
Gray Code Counterexotic1.0000
De Bruijn Sequencenumber_theory1.0000
Phyllotaxisbio1.0032
dist_from_uniform
SourceDomainValue
Aubry-André Criticalquantum0.4837
Rainfall (ORD Hourly)climate0.4830
Solar Flares Daily Peakastro0.4819
···
Gray Code Counterexotic0.0000
De Bruijn Sequencenumber_theory0.0000
Phyllotaxisbio0.0000
entropy
SourceDomainValue
Gray Code Counterexotic5.0000
De Bruijn Sequencenumber_theory5.0000
Circle Map Quasiperiodicchaos5.0000
···
Mian-Chowlanumber_theory0.0433
Aubry-André Criticalquantum0.0925
Rainfall (ORD Hourly)climate0.1593
recurrence_distance
SourceDomainValue
PID Controllerexotic0.1077
Exponential Chirpexotic0.1017
Triple Pendulummotion0.0922
···
Quartic Map (Feigenbaum)chaos0.0000
Logistic Edge-of-Chaoschaos0.0000
Logistic r=3.5 (Period-4)chaos0.0000
self_similarity
SourceDomainValue
Penrose Substitutionexotic1.0000
De Bruijn Sequencenumber_theory1.0000
Gray Code Counterexotic1.0000
···
OTOC Growthquantum0.5987
Critical Transition (Fold)chaos0.6873
Copeland-Erdősnumber_theory0.8219
transport_variability
SourceDomainValue
Exponential Chirpexotic0.2024
Critical Transition (Fold)chaos0.1606
Pulse-Width Modulationwaveform0.1334
···
Logistic r=3.2 (Period-2)chaos0.0000
Logistic r=3.5 (Period-4)chaos0.0000
Quartic Map (Feigenbaum)chaos0.0000

When It Lights Up

Wasserstein self_similarity is the distributional lens's nonstationarity detector. Signals that change character midstream — sensor drift, regime switches, concatenated recordings — score low on self_similarity while potentially scoring high on all other distributional metrics. In the atlas, Wasserstein's concentration axis separates the heavy-tailed cluster (Collatz, rainfall, forest fire) from the uniform-distribution cluster (PRNGs, De Bruijn), while self_similarity provides an orthogonal axis that catches temporal instability invisible to any single-histogram metric.

Open in Atlas
← Fisher InformationZipf–Mandelbrot (8-bit) →