Restore Branin

This file runs SMAC and then restores the branin run with an extended computation budget. This will also work for SMAC runs that have crashed and are continued.

import logging
import os

from smac.facade.smac_ac_facade import SMAC4AC
from smac.runhistory.runhistory import RunHistory
from smac.scenario.scenario import Scenario
from smac.stats.stats import Stats
from smac.utils.io.traj_logging import TrajLogger

logging.basicConfig(level=logging.INFO)


__copyright__ = "Copyright 2021, AutoML.org Freiburg-Hannover"
__license__ = "3-clause BSD"


if "__main__" == __name__:

    # Initialize scenario, using runcount_limit as budget.
    original_scenario_dict = {
        "algo": "python examples/commandline/branin.py",
        "paramfile": "examples/commandline/branin/configspace.pcs",
        "run_obj": "quality",
        "runcount_limit": 25,
        "deterministic": True,
        "output_dir": "restore_me",
    }
    original_scenario = Scenario(original_scenario_dict)

    smac = SMAC4AC(scenario=original_scenario, run_id=1)
    smac.optimize()

    print("\nBudget exhausted! Starting restoring optimization ...\n")

    # Now the output is in the folder 'restore_me/run_1' (or whatever run_id has
    # been passed to the SMAC-object above)
    old_output_dir = os.path.join(original_scenario.output_dir, "run_1")

    # We could simply modify the scenario-object, stored in
    # 'smac.solver.scenario' and start optimization again:

    # smac.solver.scenario.ta_run_limit = 50
    # smac.optimize()

    # Or, to show the whole process of recovering a SMAC-run from the output
    # directory, create a new scenario with an extended budget:
    new_scenario = Scenario(
        original_scenario_dict,
        cmd_options={
            "runcount_limit": 50,  # overwrite these args
            "output_dir": "restored",
        },
    )

    # We load the runhistory
    rh_path = os.path.join(old_output_dir, "runhistory.json")
    runhistory = RunHistory()
    runhistory.load_json(rh_path, new_scenario.cs)

    # And the stats
    stats_path = os.path.join(old_output_dir, "stats.json")
    stats = Stats(new_scenario)
    stats.load(stats_path)

    # And the trajectory
    traj_path = os.path.join(old_output_dir, "traj_aclib2.json")
    trajectory = TrajLogger.read_traj_aclib_format(fn=traj_path, cs=new_scenario.cs)
    incumbent = trajectory[-1]["incumbent"]

    # Now we can initialize SMAC with the recovered objects and restore the
    # state where we left off. By providing stats and a restore_incumbent, SMAC
    # automatically detects the intention of restoring a state.
    smac = SMAC4AC(
        scenario=new_scenario,
        runhistory=runhistory,
        stats=stats,
        restore_incumbent=incumbent,
        run_id=1,
    )

    # Because we changed the output_dir, we might want to copy the old
    # trajectory-file (runhistory and stats will be complete, but trajectory is
    # written sequentially)
    # new_traj_path = os.path.join(new_scenario.output_dir, "run_1", "traj_aclib2.json")
    # shutil.copy(traj_path, new_traj_path)

    smac.optimize()

previous

Branin

next

SPEAR-QCP with Multi-Fidelity on Instances