empyrean.SolvedCovariance

class SolvedCovariance(matrix, width, marsden_slot, dt_slot, amrat_slot, thrust_slots)[source]

Bases: object

Full tagged solved-parameter covariance from a wide OD fit. Mirrors empyrean::SolvedCovariance.

matrix is the real solved covariance, sized width × width; parameters are located by the slot fields, never by width (width 9 is Marsden-only OR one thrust segment). Canonical order is [state 6 | Marsden 3 | DT 1 | AMRAT 1 | thrust 3×k]. The Δv axes are integration-frame components (see DetermineResult.dv_frame).

Methods

__init__(matrix, width, marsden_slot, ...)

Attributes

matrix

The solved covariance, shaped (width, width).

width

Solved width (6..=17 under the current engine).

marsden_slot

Column of the first Marsden coefficient, when Marsden was solved.

dt_slot

Column of the DT scalar, when DT was solved.

amrat_slot

Column of the AMRAT scalar, when AMRAT was solved.

thrust_slots

Column triples of each fitted thrust Δv segment (one (i, i+1, i+2) per solved segment).

Parameters:
matrix: ndarray

The solved covariance, shaped (width, width).

width: int

Solved width (6..=17 under the current engine).

marsden_slot: int | None

Column of the first Marsden coefficient, when Marsden was solved.

dt_slot: int | None

Column of the DT scalar, when DT was solved.

amrat_slot: int | None

Column of the AMRAT scalar, when AMRAT was solved.

thrust_slots: list[tuple[int, int, int]]

Column triples of each fitted thrust Δv segment (one (i, i+1, i+2) per solved segment). Empty when no thrust was solved.