spectral_galerkin_heat.runner

StandaloneHeatRunner: self-contained simulation runner.

Owns the time-stepping loop, I/O and telemetry. This is the entry point for running spectral_galerkin_heat from a YAML config file.

Classes

StandaloneHeatRunner

Orchestrates the full simulation lifecycle in standalone mode.

Functions

initialize_run_logging(config, yaml_path, out_dir)

Prepare the run directory for traceability before the simulation starts.

Module Contents

spectral_galerkin_heat.runner.initialize_run_logging(config, yaml_path, out_dir)

Prepare the run directory for traceability before the simulation starts.

Copies the YAML config (and G-code file if present) into out_dir, then writes a human-readable ASCII header to logs/simulation.log. The header is written directly to the file so it sits above any timestamped log lines produced by the Python logging handlers.

Must be called after the output directory tree has been created by LocalFSIOManager.initialize().

Parameters:

  • config (Raw YAML config dict as returned by load_config().)

  • yaml_path (Absolute path to the original YAML configuration file.)

  • out_dir (Run output directory (io_manager.base_dir).)

Return type:

None

class spectral_galerkin_heat.runner.StandaloneHeatRunner(context, heat_solver, io_manager=None, config=None, yaml_path=None)

Orchestrates the full simulation lifecycle in standalone mode.

Responsibilities

  • Runs the time-stepping loop (while t < t_end).

  • Delegates periodic / end-of-run output to the LocalFSIOManager.

  • Prints ETA / telemetry to the logger.

param context:

Fully populated simulation parameters.

type context:

SimulationContext

param heat_solver:

The solver to run; build one with spectral_galerkin_heat.solvers.build_solver().

type heat_solver:

HeatSolver

param io_manager:

Output handler. Defaults to a fresh LocalFSIOManager.

type io_manager:

LocalFSIOManager, optional

run()

Executes the main simulation loop.

It initializes the logging mechanism and file outputs, builds the initial state via the solver interface, and proceeds to perform the chronological advancement loop until t_end is reached. Final states are serialized before cleanup.

Return type:

None

Parameters: