Struct IterRecord 

pub struct IterRecord {

    pub iter: i32,
    pub objective: f64,
    pub inf_pr: f64,
    pub inf_pr_internal: f64,
    pub inf_du: f64,
    pub mu: f64,
    pub d_norm: f64,
    pub regularization: f64,
    pub alpha_dual: f64,
    pub alpha_primal: f64,
    pub alpha_primal_char: char,
    pub ls_trials: i32,
    pub phase: IterPhase,
}

One row of per-iteration data — same numbers that IpoptAlgorithm prints to stdout each iteration (the “iter objective inf_pr inf_du lg(mu) ||d|| lg(rg) alpha_du alpha_pr ls” line). Captured into SolveStatistics::iterations when a JSON / programmatic consumer needs the trajectory rather than just the final state.

Field semantics mirror upstream IpOrigIterationOutput.cpp:152 (Snprintf block) so a row in JSON round-trips back into the same console table verbatim.

Fields

iter: i32

Iteration index, starting at 0.

objective: f64

Unscaled objective f(x_k) at the start of iter k.

inf_pr: f64

Primal infeasibility exactly as the console table prints it in the inf_pr column: under the default inf_pr_output=original, the unscaled constraint violation of the user’s model (for restoration rows, of the original NLP at the restoration iterate’s x). See Self::inf_pr_internal for the residual the filter and the convergence test read (gh#979).

inf_pr_internal: f64

The algorithm’s internal primal infeasibility ‖(c(x), d(x) − s)‖∞ in the scaled slack form, with the slack pushed off its bound — what the filter, the convergence test and the intermediate callback’s inf_pr see. On restoration rows it is the restoration NLP’s own residual. Differs from Self::inf_pr whenever a slack sits away from d(x) or scaling is active.

inf_du: f64

Dual infeasibility (max-norm of grad-Lagrangian).

mu: f64

Barrier parameter μ.

d_norm: f64

||d_xs||_∞ of the search step. 0.0 on iter 0 (no step yet).

regularization: f64

Hessian regularization δ_w applied this iter; 0.0 when no regularization was needed (printed as - in the console).

alpha_dual: f64

Dual step length.

alpha_primal: f64

Primal step length.

alpha_primal_char: char

Single-character tag for the alpha-primal column (f, h, r for restoration etc.) — matches upstream’s per-iter tag.

ls_trials: i32

Number of backtracking line-search trials this iter.

phase: IterPhase

Which phase printed this row: the main IPM, or the restoration sub-solve (the console’s r-suffixed rows). A restoration row’s iter is the restoration solver’s counter, which continues the outer count, so the outer row that leaves restoration repeats the last restoration row’s index — exactly as the console does.

Trait Implementations

impl Clone for IterRecord

fn clone(&self) -> IterRecord

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for IterRecord

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

Formats the value using the given formatter.

impl Default for IterRecord

fn default() -> IterRecord

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for IterRecord

fn deserialize<__D>( __deserializer: __D, ) -> Result<IterRecord, <__D as Deserializer<'de>>::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for IterRecord

fn serialize<__S>( &self, __serializer: __S, ) -> Result<<__S as Serializer>::Ok, <__S as Serializer>::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.