Reference library
A practical method for calculating largest Lyapunov exponents from small data sets
Paper summary
A delay-reconstruction estimator tracks nearby trajectory segments with temporal-neighbor exclusion and fits a slope to average logarithmic divergence. Tests explore noise, embedding and quasiperiodic/stochastic controls. A bounded trajectory produces eventual distance saturation; short portions of nonchaotic curves may resemble positive scaling regions. The method supports conditional evidence with diagnostic plots rather than blind reliance on a fitted slope.
Relevance to the project
Assistant interpretation: useful orientation for historical scalar traces when complete model state cannot be recovered. For known neural code, paired-state interventions and a declared transition/state are preferable to assuming that one movement score embeds a high-dimensional plastic system.
Targeted manuscript reading, not all tables/figures or parameter robustness results audited. Positive finite-time separation alone is insufficient to rule out noise, transients and nonstationarity; the paper's optimistic detection language must be read with later surrogate-testing cautions.
Source record
- Reference ID
- rosenstein-lyapunov-1993
- Archived version
- Manuscript dated20Nov1992,60PDFpages, associated with PhysicaD65:117–134(1993)
- Summary provenance
- reference-library/papers/rosenstein-lyapunov-1993/metadata.json · summary
- Review status
- source-scoped reading; not replication
- Reading scope
- Abstract and PDF pages 7–10,13–15,18–20 read. These are PDF positions in the 60-page manuscript, not journal pages. Remaining sections, tables and separate figure pages not fully read.