SegmentInformation

class ell1fit.information.SegmentInformation(name: str, raw: ndarray, effective: ndarray, coupling: ndarray, nuisance: ndarray)[source]

Bases: object

One observation’s information about the globally shared parameters.

Attributes:
namestr

Identifier for the segment, used in tables and error messages.

rawnp.ndarray

The shared block \(A_i\) on its own: what this segment would say about the shared parameters if its spin parameters were known exactly.

effectivenp.ndarray

The Schur complement \(A_i - B_i D_i^{-1} B_i^{T}\): what survives once the segment’s own spin parameters have been fitted away. This is the quantity that adds across segments.

couplingnp.ndarray

The off-diagonal block \(B_i\), kept so the full arrow matrix can be reassembled.

nuisancenp.ndarray

The segment’s own spin block \(D_i\).

Attributes Summary

coupling

dilution

Fraction of each shared parameter's information surviving the local fit.

effective

n_nuisance

Number of per-segment spin parameters profiled away.

n_shared

Number of globally shared parameters.

name

nuisance

raw

Attributes Documentation

coupling: ndarray = <dataclasses._MISSING_TYPE object>
dilution

Fraction of each shared parameter’s information surviving the local fit.

One means the segment’s own spin parameters take nothing; a number near zero means the local fit can very nearly reproduce the orbital signature by itself, so the segment is close to useless for that parameter however many photons it holds.

effective: ndarray = <dataclasses._MISSING_TYPE object>
n_nuisance

Number of per-segment spin parameters profiled away.

n_shared

Number of globally shared parameters.

name: str = <dataclasses._MISSING_TYPE object>
nuisance: ndarray = <dataclasses._MISSING_TYPE object>
raw: ndarray = <dataclasses._MISSING_TYPE object>