Struct PauliObservable 

pub struct PauliObservable { /* private fields */ }

Weighted sum of joint Pauli observables, H = sum_k c_k P_k.

Terms are kept canonical: factors sorted by qubit, identical Pauli strings merged by summing coefficients, terms ordered by string. An empty factor list is the identity and contributes its coefficient as a constant offset. The qubit-wise-commuting grouping the grouped evaluation route uses is computed lazily and cached; mutation invalidates the cache.

Implementations

impl PauliObservable

pub fn new() -> Self

pub fn from_terms( terms: impl IntoIterator<Item = (f64, Vec<PauliTerm>)>, ) -> Result<Self>

Build from (coefficient, factors) pairs, the Hamiltonian shape Simulate::expectation_gradient takes.

pub fn add_term( &mut self, coefficient: f64, factors: Vec<PauliTerm>, ) -> Result<()>

Add coefficient times the Pauli string factors, merging into an existing term with the same string.

Errors

Rejects a non-finite coefficient and duplicate factors on one qubit.

pub fn terms(&self) -> &[(f64, Vec<PauliTerm>)]

Canonical (coefficient, factors) pairs, ordered by Pauli string.

pub fn num_terms(&self) -> usize

pub fn num_groups(&self) -> usize

Number of qubit-wise-commuting groups, computing the grouping if needed. Identity terms belong to no group.

pub fn split_identity(&self) -> (f64, PauliObservable)

The constant term’s coefficient and the rest of the sum.

Var(H + cI) = Var(H), so a variance squares the traceless part rather than the whole sum: at a large c the constant dominates both <H^2> and <H>^2 and the difference loses the spread it was meant to report.

pub fn square(&self) -> PauliObservable

H^2 as a Pauli sum, the second moment Simulate::observable_variance reads Var(H) = <H^2> - <H>^2 from.

Every coefficient of the square is real. Two Pauli strings either commute, and their product carries no phase, or anticommute, and the (j, k) and (k, j) products carry opposite imaginary phases that cancel. Phases are tracked as powers of i so that cancellation is exact rather than a subtraction of two nearly equal floats.

Costs T^2 string products over T terms, so it suits the tensor products and small Hermitian matrices an observable request names rather than a molecular Hamiltonian.

Trait Implementations

impl Add for PauliObservable

type Output = PauliObservable

The resulting type after applying the + operator.

fn add(self, rhs: PauliObservable) -> PauliObservable

Performs the + operation.

impl Clone for PauliObservable

fn clone(&self) -> PauliObservable

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for PauliObservable

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for PauliObservable

fn default() -> PauliObservable

Returns the “default value” for a type.

impl Mul<f64> for PauliObservable

type Output = PauliObservable

The resulting type after applying the * operator.

fn mul(self, rhs: f64) -> PauliObservable

Performs the * operation.

impl Neg for PauliObservable

type Output = PauliObservable

The resulting type after applying the - operator.

fn neg(self) -> PauliObservable

Performs the unary - operation.

impl Sub for PauliObservable

type Output = PauliObservable

The resulting type after applying the - operator.

fn sub(self, rhs: PauliObservable) -> PauliObservable

Performs the - operation.