Brownian Walk

stochastic-process · 37 views
stochastic-processstochastic

What It Is

Cumulative sum of Gaussian increments --- the canonical random walk, self-affine with Hurst exponent H=0.5 and PSD ~ 1/f²

Interpretation

Standard analysis sees: smooth / autocorrelated; long-range memory (persistent); volatility-clustering (bursty); nonstationary / drifting. The atlas additionally detects Nonstationarity:adf_pvalue.

What standard analysis sees
tail heaviness0.39
asymmetry0.15
occupancy0.78
short-range corr0.88
long-range memory0.90
spectral colour0.25
periodicity0.72
complexity0.67
time-irreversibility0.38
volatility clustering0.87
multifractality0.39
dimensionality0.19
nonstationarity0.87
What the atlas adds
Nonstationarity:adf_pvalue+2.7z
unit-root nonstationarity (ADF cannot reject random-walk null)
Atlas-extreme metrics the standard bank can’t predict for this source
Dodecagonal (Stampfli):pisot_triplet_coherence+3.3zbank-miss 1.2σ

Composition

dtypefloat64
range[-111.9, 132.9]
unique values16384 / 16384
mean ± std14.5 ± 56.1

Render Gallery

Atlas Position

Nearest neighborDistance
Geometric Brownian Motion2.29
Perlin Noise2.29
Regime Switching2.58

Open in Atlas →

Which Geometries Light Up

Catch24Catch24:CO_f1ecacrank 4/3072334.6222
Catch24Catch24:DN_OutlierInclude_n_001_mdrmdrank 4/3070.3049
in stochastic-process
alphabetical
← / → within domain · ⇧← / ⇧→ alphabetical · ⇧← / ⇧→ inside an open render = same view across sources