porcupy.utils.interactive_visualization

Interactive visualization module for the Crested Porcupine Optimizer (CPO).

This module provides interactive visualization tools and dashboard components for monitoring and analyzing the CPO algorithm’s performance in real-time.

Classes

OptimizationDashboard	Interactive dashboard for monitoring CPO optimization in real-time.
ParameterTuningDashboard	Interactive dashboard for parameter tuning and sensitivity analysis.

Functions

create_interactive_optimization_plot(objective_func, ...)

Create an interactive plot for exploring the optimization landscape.

Module Contents

class porcupy.utils.interactive_visualization.OptimizationDashboard(objective_func: Callable, bounds: Tuple[numpy.ndarray, numpy.ndarray], dimensions: int = 2, update_interval: float = 0.5, figsize: Tuple[int, int] = (15, 10))

Interactive dashboard for monitoring CPO optimization in real-time.

This dashboard displays multiple visualizations simultaneously, including: - Convergence history - Population size history - Diversity metrics - Defense mechanism usage - Current porcupine positions

The dashboard updates in real-time during optimization.

objective_func

bounds

dimensions = 2

update_interval = 0.5

figsize = (15, 10)

iterations = []

best_costs = []

pop_sizes = []

diversity_history = []

defense_counts

position_history = []

best_position_history = []

is_running = False

current_iteration = 0

update(iteration: int, best_cost: float, pop_size: int, positions: numpy.ndarray, best_position: numpy.ndarray, defense_types: List[str])

Update the dashboard with new optimization data.

Parameters:

• iteration (int) – Current iteration number.

• best_cost (float) – Current best cost value.

• pop_size (int) – Current population size.

• positions (ndarray) – Current positions of the porcupines, shape (pop_size, dimensions).

• best_position (ndarray) – Current global best position, shape (dimensions,).

• defense_types (list) – List of defense mechanisms used by each porcupine.

start_monitoring()

Start the dashboard monitoring thread.

stop_monitoring()

Stop the dashboard monitoring thread.

save_dashboard(save_path: str, dpi: int = 300)

Save the current dashboard state as an image.

Parameters:

• save_path (str) – Path to save the dashboard image.

• dpi (int, optional) – DPI for the saved image (default: 300).

close()

Close the dashboard and clean up resources.

class porcupy.utils.interactive_visualization.ParameterTuningDashboard(parameter_name: str, parameter_range: List[float], result_metric: str = 'Best Cost', figsize: Tuple[int, int] = (12, 8))

Interactive dashboard for parameter tuning and sensitivity analysis.

This dashboard allows for real-time visualization of how different parameter values affect the performance of the CPO algorithm.

parameter_name

parameter_range

result_metric = 'Best Cost'

figsize = (12, 8)

results = []

update(parameter_value: float, result: float, convergence_history: List[float] | None = None)

Update the dashboard with new parameter tuning results.

Parameters:

• parameter_value (float) – Value of the parameter being tested.

• result (float) – Result metric value (e.g., best cost).

• convergence_history (list, optional) – Convergence history for this parameter value.

save_dashboard(save_path: str, dpi: int = 300)

Save the current dashboard state as an image.

Parameters:

• save_path (str) – Path to save the dashboard image.

• dpi (int, optional) – DPI for the saved image (default: 300).

close()

Close the dashboard and clean up resources.

porcupy.utils.interactive_visualization.create_interactive_optimization_plot(objective_func: Callable, bounds: Tuple[numpy.ndarray, numpy.ndarray], initial_positions: numpy.ndarray, figsize: Tuple[int, int] = (10, 8))

Create an interactive plot for exploring the optimization landscape.

Parameters:

• objective_func (callable) – The objective function to visualize.

• bounds (tuple) – A tuple (lb, ub) containing the lower and upper bounds.

• initial_positions (ndarray) – Initial positions of the porcupines, shape (pop_size, 2).

• figsize (tuple, optional) – Figure size as (width, height) in inches (default: (10, 8)).

Returns:

A tuple containing (fig, ax, scatter) for further customization.

Return type:

tuple