interp_poly opsignal × signal → tableimport fullseye as fs; fs.ledger.poly_fit(x, y, degree) (実装を直接呼ぶなら import mathops; mathops.poly_fit(x, y, degree)、台帳から引くなら opsmath.get("poly_fit"))Least-squares polynomial fit with its conditioning on the record.
Fits y ≈ c[0] x^d + ... + c[d] (coefficients highest-power-first, the
:func:poly_eval / np.polyval convention) by SVD least squares on the
Vandermonde matrix. degree must be an integer >= 0 with at least
degree + 1 samples (fail-closed: an exactly-determined fit is allowed,
an under-determined one is not).
Returns a dict — the fit and its health, inseparable:
coeffs (degree + 1,) float64 · degree · cond the Vandermonde
condition number (:func:mat_cond of the design matrix) · rms_residual
root-mean-square of y - p(x).
The conditioning mechanism: when cond > POLY_COND_WARN (1e10) a
RuntimeWarning is emitted and the number is in the result — an
equispaced degree-10 fit on raw pixel coordinates is already past it. High
degree on a raw coordinate range is the classic double trap: the
Vandermonde columns become near-collinear (digits lost, coefficients
unstable) and the fit oscillates between nodes (Runge phenomenon, Runge
1901). Centre and scale x to [-1, 1] first, or keep degree ≤ ~6.
HALCON: no public polynomial-fitting tuple operator (fitting of this kind lives inside HALCON’s calibration internals).
mathops の全 op は入力を検証してから計算する(黙って通さない):
ValueError — float64 への強制変換は虚部を黙って捨てる(numpy は ComplexWarning だけ出して「もっともらしく間違った」実数を返す)。.real/.imag/abs() を明示するか、複素対応の complexops を使う。ValueError — マスクを剥がして下の生値を使う暗黙変換を拒否。埋める/落とすを明示する。ValueError(件数を明示して拒否 — 結果全体に伝播するため)。ValueError。reshape を明示する)。stat_histogram の bins は mathops.MAX_ELEMENTS(2^26 ≈ 6700 万要素)超で ValueError。py -3.11 examples/math_metrology.pypy -3.11 examples/poc_strain_history.pypy -3.11 examples/poc_thermal_radiometry.pypy -3.11 examples/signal_filter.pytable を入力に取れる)—
interp_poly)interp_linear · interp_cubic · interp_scattered · poly_eval · poly_roots
Provenance: mathops.py — MATH operator registry. この per-op ノートは tools/opdocs.py md が自動生成(手編集しない)。
© 2026 Kazufumi Furuse — Fullseye operator documentation. Licensed under Apache-2.0.