Datasets:
Formats:
csv
Languages:
English
Size:
1K - 10K
Tags:
mathematics
inverse-problems
observability
noncommutative-algebra
symbolic-computation
calibration-rigidity
|
Download docs/API_REFERENCE.md from PureOne/AELYRION-NOULITH-III: direct link, hf CLI and curl.
- Browser
- Download file 2.36 kB
-
https://huggingface.co/datasets/PureOne/AELYRION-NOULITH-III/resolve/main/docs/API_REFERENCE.md
- Command line
-
hf download hf://datasets/PureOne/AELYRION-NOULITH-III/docs/API_REFERENCE.md
-
curl -L -o API_REFERENCE.md https://huggingface.co/datasets/PureOne/AELYRION-NOULITH-III/resolve/main/docs/API_REFERENCE.md
2.36 kB
Exact reference API
Install the included src/noulith/ package with python -m pip install -e ..
The functions below are existing upstream entry points.
Function in noulith.third |
Meaning / boundary |
|---|---|
all_components(m) |
Star, triangle, and twin descriptors; theorem domain m>=6 |
component_equations(m, component) |
Exact integer matrix in ascending logarithmic edge order |
intersection_dimension(m, components) |
Dimension computed from stacked exact equations |
component_membership(m, q) |
Components containing a coefficient point under the implemented relations |
certify_scalar(m, q) |
Exact rational certificate, m>=6 and abs(q+1)<1/20 |
predicted_rank(m, q) |
Third-deficit rank from classification conditions |
direct_block_rank(m, q, r) |
Independent all-subset calculation; exponential in m |
optimal_calibration_subfamily(m) |
Attaining selection of twin outlier pairs, m>=6 |
calibration_classes(m, selected_twins) |
Equivalence classes of edge scales constrained by the selected twin subspaces |
example_point(m, kind, permutation=None) |
Nearby rational example point |
q_from_exponents(exponents, base=Rational(1001,1000)) |
Rational coefficients from directed antisymmetric integer exponents |
Named examples: origin, star, triangle, twin, uniform_star, star_twin,
edge, double_star, generic.
Verify a raw witness
from noulith.algebra import TwistedAlgebra
from noulith.third import example_point, certify_scalar
m = 6
q = example_point(m, "generic")
certificate = certify_scalar(m, q)
assert certificate.visible
blocks = certificate.full_partition(m)
algebra = TwistedAlgebra(m, q)
raw = algebra.isolated_response(blocks)
assert len(blocks) == m - 3
assert raw == algebra.isolated_unit(blocks) * certificate.normalized_response
assert raw != 0
This checks the actual sparse algebra, including its crossing factor. It should not be replaced by a rounded near-zero test.
Scope
The local analytic theorem and strict real-rational implementation neighborhood are distinct. Passing an arbitrary dictionary to a low-level routine does not extend a theorem. A selected calibration pair represents an entire twin-subspace constraint, not a detector sample. Do not apply the third-deficit theorem at m=5 or the generic core formula at r=0.