dynsys_lyapunov_spectrum — MATH dynsys op

• Data kinds: none → signal (an op determined by its arguments alone — it takes no image or data input)

• Call: import fullseye as fs; fs.ledger.dynsys_lyapunov_spectrum(system='lorenz', params=None, x0=None, t_end=200.0, dt=0.005, burn_in=20.0) (to call the implementation directly, import mathops; mathops.dynsys_lyapunov_spectrum(system='lorenz', params=None, x0=None, t_end=200.0, dt=0.005, burn_in=20.0); from the registry, opsmath.get("dynsys_lyapunov_spectrum"))

Usage

The Lyapunov spectrum by tangent flow + QR — and the sum you can check.

Integrates the state together with an orthonormal frame of tangent vectors

(the variational equation `dY/dt = J(x) Y`), re-orthonormalising by QR at

every step and accumulating `log` of the diagonal. The exponents come out

ordered, largest first.

★★Why this earns its place — the trace identity. The sum of the exponents

equals the time-average of the divergence of the field:

sum(lambda_i) == <div f>

For Lorenz the divergence is the constant `-(sigma + 1 + beta)`, so the

sum is known in closed form: `-13.6667` for the classical parameters. That

is an exact target the attractor picture cannot provide. The published largest

exponent (≈ 0.906 for sigma=10, beta=8/3, rho=28) is a second, independent

check.

Returns a `signal`: the exponents, descending.

Raises `ValueError: unknown system; non-finite input; burn_in` not

shorter than `t_end`; a trajectory that left float range.

Limits: the exponents converge like `1/sqrt(T)` — a short window gives a

plausible but wrong spectrum. The trace identity converges much faster and is

the honest gate; the individual exponents need long windows.

HALCON: no operator.

Family-wide input contract (fail-closed)

Every mathops op validates its input before computing (nothing slips through silently):

• **complex input raises ValueError** — coercing to float64 silently discards the imaginary part (numpy only emits a ComplexWarning and returns a plausible-looking wrong real number). State .real/.imag/abs() explicitly, or use complexops, which handles complex data.

• **masked arrays with masked elements raise ValueError** — the implicit conversion that peels off the mask and uses the raw values underneath is refused. Say explicitly whether to fill or to drop.

• **NaN/Inf raises ValueError on every input** (refused with the count stated — it propagates through the whole result).

• Shapes are strict: 1-D and 2-D are never implicitly promoted or broadcast (a matrix in a vector slot, or a vector in a matrix slot, raises ValueError; reshape explicitly).

• Size cap: ops that take a matrix, and the stat_histogram bins, raise ValueError beyond mathops.MAX_ELEMENTS (2^26 ≈ 67 million elements).

Detailed usage guide

• math_metrology family guide

References (sample data, literature)

• Sample-data catalog (download URLs / licences) — 2-D uses skimage.data (BSD/public domain) plus synthetic images; 3-D lists download URLs for real data sources (Stanford, PDS, …).

• Operator provenance and references — the sources of the research/methods this op family came from.

• The canonical algorithm (author, year) and its uses are named in the family usage guide above.

Runnable examples (verified samples that actually call this op)

• poc_what_a_picture_cannot_check — py -3.11 examples/poc_what_a_picture_cannot_check.py

Ops the type connects to (they accept signal as input)

mat_solve · mat_lstsq · stat_describe · stat_histogram · stat_zscore · interp_linear · interp_cubic · interp_scattered

Same category (dynsys)

ode_flow_states · ode_vector_field_grid · dynsys_poincare_section · dynsys_bifurcation_map · dynsys_correlation_dimension


*Provenance: mathops.py — MATH operator registry. This per-op note is generated by tools/opdocs.py md (do not hand-edit).*

© 2026 Kazufumi Furuse — Fullseye operator documentation. Licensed under Apache-2.0.