Struct OptimizationVisualizer 

pub struct OptimizationVisualizer { /* private fields */ }

Optimization visualizer for real-time debugging with Rerun.

This struct manages all visualization state and provides methods for logging various aspects of the optimization process.

Implementations

impl OptimizationVisualizer

pub fn new(enabled: bool) -> Result<Self, Box<dyn Error>>

Create a new optimization visualizer.

Arguments

• enabled - Whether to enable visualization

Returns

A new visualizer instance.

pub fn new_with_options( enabled: bool, save_path: Option<&str>, ) -> Result<Self, Box<dyn Error>>

Create a new optimization visualizer with file save option.

Arguments

• enabled - Whether to enable visualization
• save_path - Optional path to save recording to file instead of spawning viewer

pub fn is_enabled(&self) -> bool

Check if visualization is enabled and active.

pub fn log_scalars( &self, iteration: usize, cost: f64, gradient_norm: f64, damping: f64, step_norm: f64, step_quality: Option<f64>, ) -> Result<(), Box<dyn Error>>

Log scalar values for time series plots.

Each metric is logged to a separate entity path for independent scaling.

Arguments

• iteration - Current iteration number
• cost - Current cost value
• gradient_norm - L2 norm of the gradient
• damping - Current damping parameter (λ)
• step_norm - L2 norm of the parameter update
• step_quality - Step quality metric (ρ)

pub fn log_timing( &self, iteration: usize, duration_ms: f64, ) -> Result<(), Box<dyn Error>>

Log iteration timing information.

Arguments

• iteration - Current iteration number
• duration_ms - Time taken for this iteration in milliseconds

pub fn log_hessian( &self, hessian: Option<&SparseColMat<usize, f64>>, iteration: usize, ) -> Result<(), Box<dyn Error>>

Log sparse Hessian matrix as a heat map image.

The Hessian is downsampled to a fixed 100×100 image for visualization. Uses a blue-white-red color scheme:

• Negative values: blue
• Zero values: white
• Positive values: red

Arguments

• hessian - The sparse Hessian matrix
• iteration - Current iteration number

pub fn log_gradient( &self, gradient: Option<&Mat<f64>>, iteration: usize, ) -> Result<(), Box<dyn Error>>

Log gradient vector as a tensor visualization.

The gradient is downsampled to a fixed 100-element vector and displayed as a horizontal bar image.

Arguments

• gradient - The gradient vector
• iteration - Current iteration number

pub fn log_manifolds( &self, variables: &HashMap<String, VariableEnum>, iteration: usize, ) -> Result<(), Box<dyn Error>>

Log current manifold states for SE2/SE3 problems.

This visualizes the current pose estimates during optimization. Only the latest poses are shown (previous iterations are cleared).

Arguments

• variables - Current variable values
• iteration - Current iteration number

pub fn log_convergence(&self, status: &str) -> Result<(), Box<dyn Error>>

Log convergence status and final summary.

Arguments

• status - Convergence status message

pub fn log_initial_graph( &self, graph: &Graph, scale: f32, ) -> Result<(), Box<dyn Error>>

Log the initial graph structure before optimization.

This visualizes only the initial poses (vertices), not edges.

Arguments

• graph - The graph structure loaded from G2O file
• scale - Scale factor for visualization

Trait Implementations

impl Debug for OptimizationVisualizer

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

Formats the value using the given formatter.

impl Default for OptimizationVisualizer

fn default() -> Self

Returns the “default value” for a type.