file

get_experiment_filename(PREFIX)

Returns the name of the experiment.

Parameters:

Name	Type	Description	Default
PREFIX	str	Prefix of the experiment.	required

Returns:

Name	Type	Description
filename	str	Name of the experiment.

Examples:

>>> from spotpython.utils.file import get_experiment_name
>>> from spotpython.utils.init import fun_control_init
>>> fun_control = fun_control_init(PREFIX="branin")
>>> PREFIX = fun_control["PREFIX"]
>>> filename = get_experiment_filename(PREFIX)

Source code in spotpython/utils/file.py

def get_experiment_filename(PREFIX):
    """Returns the name of the experiment.

    Args:
        PREFIX (str): Prefix of the experiment.

    Returns:
        filename (str): Name of the experiment.

    Examples:
        >>> from spotpython.utils.file import get_experiment_name
        >>> from spotpython.utils.init import fun_control_init
        >>> fun_control = fun_control_init(PREFIX="branin")
        >>> PREFIX = fun_control["PREFIX"]
        >>> filename = get_experiment_filename(PREFIX)
    """
    if PREFIX is None:
        return None
    else:
        filename = "spot_" + PREFIX + "_experiment.pickle"
    return filename

get_experiment_from_PREFIX(PREFIX, return_dict=True)

Setup the experiment based on the PREFIX provided and return the relevant configuration and control objects.

Parameters:

Name	Type	Description	Default
PREFIX	str	The prefix for the experiment filename.	required
return_dict	bool	Whether to return the configuration and control objects as a dictionary. If False, a tuple is returned: “(config, fun_control, design_control, surrogate_control, optimizer_control).” Defaults to True.	True

Returns:

Name	Type	Description
dict	dict	Dictionary containing the configuration and control objects.

Example

from spotpython.utils.file import get_experiment_from_PREFIX config = get_experiment_from_PREFIX(“100”)[“config”]

Source code in spotpython/utils/file.py

def get_experiment_from_PREFIX(PREFIX, return_dict=True) -> dict:
    """
    Setup the experiment based on the PREFIX provided and return the relevant configuration
    and control objects.

    Args:
        PREFIX (str):
            The prefix for the experiment filename.
        return_dict (bool, optional):
            Whether to return the configuration and control objects as a dictionary.
            If False, a tuple is returned:
            "(config, fun_control, design_control, surrogate_control, optimizer_control)."
            Defaults to True.

    Returns:
        dict: Dictionary containing the configuration and control objects.

    Example:
        >>> from spotpython.utils.file import get_experiment_from_PREFIX
        >>> config = get_experiment_from_PREFIX("100")["config"]

    """
    experiment_name = get_experiment_filename(PREFIX)
    spot_tuner, fun_control, design_control, surrogate_control, optimizer_control = load_experiment(experiment_name)
    config = get_tuned_architecture(spot_tuner, fun_control)
    if return_dict:
        return {
            "config": config,
            "fun_control": fun_control,
            "design_control": design_control,
            "surrogate_control": surrogate_control,
            "optimizer_control": optimizer_control,
        }
    else:
        return config, fun_control, design_control, surrogate_control, optimizer_control

load_and_run_spot_python_experiment(spot_pkl_name)

Loads and runs a spot experiment.

Parameters:

Name	Type	Description	Default
spot_pkl_name	str	The name of the spot experiment file.	required

Returns:

Name	Type	Description
tuple	tuple	A tuple containing the spot tuner, fun control, design control, surrogate control, optimizer control, and the tensorboard process object (p_popen).

Notes

p_open is deprecated and should be removed in future versions. It returns None.

Examples:

>>> from spotpython.utils.file import load_and_run_spot_python_experiment
>>> spot_tuner = load_and_run_spot_python_experiment("spot_branin_experiment.pickle")

Source code in spotpython/utils/file.py

def load_and_run_spot_python_experiment(spot_pkl_name) -> tuple:
    """Loads and runs a spot experiment.

    Args:
        spot_pkl_name (str):
            The name of the spot experiment file.

    Returns:
        tuple: A tuple containing the spot tuner, fun control,
               design control, surrogate control, optimizer control,
               and the tensorboard process object (p_popen).

    Notes:
        p_open is deprecated and should be removed in future versions.
        It returns None.

    Examples:
        >>> from spotpython.utils.file import load_and_run_spot_python_experiment
        >>> spot_tuner = load_and_run_spot_python_experiment("spot_branin_experiment.pickle")

    """
    p_open = None
    (spot_tuner, fun_control, design_control, surrogate_control, optimizer_control) = load_experiment(spot_pkl_name)
    print("\nLoaded fun_control in spotRun():")
    # pprint.pprint(fun_control)
    print(gen_design_table(fun_control))
    setup_paths(fun_control["TENSORBOARD_CLEAN"])
    spot_tuner.init_spot_writer()
    # if fun_control["tensorboard_start"]:
    #     p_open = start_tensorboard()
    # else:
    #     p_open = None
    spot_tuner.run()
    # # tensorboard --logdir="runs/"
    # stop_tensorboard(p_open)
    print(gen_design_table(fun_control=fun_control, spot=spot_tuner))
    return spot_tuner, fun_control, design_control, surrogate_control, optimizer_control, p_open

load_cifar10_data(data_dir='./data')

Loads the CIFAR10 dataset.

Parameters:

Name	Type	Description	Default
data_dir	str	Directory to save the data. Defaults to “./data”.	'./data'

Returns:

Name	Type	Description
trainset	CIFAR10	Training dataset.

Examples:

>>> from spotpython.utils.file import load_cifar10_data
>>> trainset = load_cifar10_data(data_dir="./data")

Source code in spotpython/utils/file.py

def load_cifar10_data(data_dir="./data"):
    """Loads the CIFAR10 dataset.

    Args:
        data_dir (str, optional): Directory to save the data. Defaults to "./data".

    Returns:
        trainset (torchvision.datasets.CIFAR10): Training dataset.

    Examples:
        >>> from spotpython.utils.file import load_cifar10_data
        >>> trainset = load_cifar10_data(data_dir="./data")

    """
    transform = transforms.Compose([transforms.ToTensor(), transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])
    trainset = torchvision.datasets.CIFAR10(root=data_dir, train=True, download=True, transform=transform)
    testset = torchvision.datasets.CIFAR10(root=data_dir, train=False, download=True, transform=transform)
    return trainset, testset

load_core_model_from_file(coremodel, dirname='userModel')

Loads a core model from a python file.

Parameters:

Name	Type	Description	Default
coremodel	str	Name of the core model.	required
dirname	str	Directory name. Defaults to “userModel”.	'userModel'

Returns:

Name	Type	Description
coremodel	object	Core model.

Source code in spotpython/utils/file.py

def load_core_model_from_file(coremodel, dirname="userModel"):
    """Loads a core model from a python file.

    Args:
        coremodel (str): Name of the core model.
        dirname (str, optional): Directory name. Defaults to "userModel".

    Returns:
        coremodel (object): Core model.

    """
    sys.path.insert(0, "./" + dirname)
    module = importlib.import_module(coremodel)
    core_model = getattr(module, coremodel)
    return core_model

load_dict_from_file(coremodel, dirname='userModel')

Loads a dictionary from a json file.

Parameters:

Name	Type	Description	Default
coremodel	str	Name of the core model.	required
dirname	str	Directory name. Defaults to “userModel”.	'userModel'

Returns:

Name	Type	Description
dict	dict	Dictionary with the core model.

Source code in spotpython/utils/file.py

def load_dict_from_file(coremodel, dirname="userModel"):
    """Loads a dictionary from a json file.

    Args:
        coremodel (str): Name of the core model.
        dirname (str, optional): Directory name. Defaults to "userModel".

    Returns:
        dict (dict): Dictionary with the core model.

    """
    file_path = os.path.join(dirname, f"{coremodel}.json")
    if os.path.isfile(file_path):
        with open(file_path, "r") as f:
            dict_tmp = json.load(f)
            dict = dict_tmp[coremodel]
    else:
        print(f"The file {file_path} does not exist.")
        dict = None
    return dict

load_experiment(PKL_NAME=None, PREFIX=None)

Loads the experiment from a pickle file. If PKL_NAME is None and PREFIX is not None, the experiment is loaded based on the PREFIX using the get_experiment_filename function. If the spot tuner object and the fun control dictionary do not exist, an error is thrown. If the design control, surrogate control, and optimizer control dictionaries do not exist, a warning is issued and None is assigned to the corresponding variables.

Parameters:

Name	Type	Description	Default
PKL_NAME	str	Name of the pickle file. Defaults to None.	None
PREFIX	str	Prefix of the experiment. Defaults to None.	None

Returns:

Name	Type	Description
spot_tuner	object	The spot tuner object.
fun_control	dict	The function control dictionary.
design_control	dict	The design control dictionary.
surrogate_control	dict	The surrogate control dictionary.
optimizer_control	dict	The optimizer control dictionary.

Notes

The corresponding save_experiment function is part of the class spot.

Examples:

>>> from spotpython.utils.file import load_experiment
>>> spot_tuner, fun_control, design_control, _, _ = load_experiment("spot_0_experiment.pickle")

Source code in spotpython/utils/file.py

def load_experiment(PKL_NAME=None, PREFIX=None):
    """
    Loads the experiment from a pickle file.
    If PKL_NAME is None and PREFIX is not None, the experiment is loaded based on the PREFIX
    using the get_experiment_filename function.
    If the spot tuner object and the fun control dictionary do not exist, an error is thrown.
    If the design control, surrogate control, and optimizer control dictionaries do not exist, a warning is issued
    and `None` is assigned to the corresponding variables.

    Args:
        PKL_NAME (str): Name of the pickle file. Defaults to None.
        PREFIX (str, optional): Prefix of the experiment. Defaults to None.

    Returns:
        spot_tuner (object): The spot tuner object.
        fun_control (dict): The function control dictionary.
        design_control (dict): The design control dictionary.
        surrogate_control (dict): The surrogate control dictionary.
        optimizer_control (dict): The optimizer control dictionary.

    Notes:
        The corresponding save_experiment function is part of the class spot.

    Examples:
        >>> from spotpython.utils.file import load_experiment
        >>> spot_tuner, fun_control, design_control, _, _ = load_experiment("spot_0_experiment.pickle")

    """
    if PKL_NAME is None and PREFIX is not None:
        PKL_NAME = get_experiment_filename(PREFIX)
    with open(PKL_NAME, "rb") as handle:
        experiment = pickle.load(handle)
        print(f"Loaded experiment from {PKL_NAME}")
    # assign spot_tuner and fun_control only if they exist otherwise throw an error
    if "spot_tuner" not in experiment:
        raise ValueError("The spot tuner object does not exist in the pickle file.")
    if "fun_control" not in experiment:
        raise ValueError("The fun control dictionary does not exist in the pickle file.")
    spot_tuner = experiment["spot_tuner"]
    fun_control = experiment["fun_control"]
    # assign the rest of the dictionaries if they exist otherwise assign None
    if "design_control" not in experiment:
        design_control = None
        # issue a warning
        print("The design control dictionary does not exist in the pickle file. Returning None.")
    else:
        design_control = experiment["design_control"]
    if "surrogate_control" not in experiment:
        surrogate_control = None
        # issue a warning
        print("The surrogate control dictionary does not exist in the pickle file. Returning None.")
    else:
        surrogate_control = experiment["surrogate_control"]
    if "optimizer_control" not in experiment:
        # issue a warning
        print("The optimizer control dictionary does not exist in the pickle file. Returning None.")
        optimizer_control = None
    else:
        optimizer_control = experiment["optimizer_control"]
    return spot_tuner, fun_control, design_control, surrogate_control, optimizer_control

load_pickle(filename)

Loads a pickle file. Add .pkl to the filename.

Parameters:

Name	Type	Description	Default
filename	str	Name of the pickle file.	required

Returns:

Type	Description
object	Loaded object.

Examples:

>>> from spotpython.utils.file import load_pickle
>>> obj = load_pickle(filename="obj.pkl")

Source code in spotpython/utils/file.py

def load_pickle(filename: str):
    """Loads a pickle file.
        Add .pkl to the filename.

    Args:
        filename (str): Name of the pickle file.

    Returns:
        (object): Loaded object.

    Examples:
        >>> from spotpython.utils.file import load_pickle
        >>> obj = load_pickle(filename="obj.pkl")
    """
    filename = filename + ".pkl"
    with open(filename, "rb") as f:
        obj = pickle.load(f)
    return obj

save_pickle(obj, filename)

Saves an object as a pickle file. Add .pkl to the filename.

Parameters:

Name	Type	Description	Default
obj	object	Object to be saved.	required
filename	str	Name of the pickle file.	required

Examples:

>>> from spotpython.utils.file import save_pickle
>>> save_pickle(obj, filename="obj.pkl")

Source code in spotpython/utils/file.py

def save_pickle(obj, filename: str):
    """Saves an object as a pickle file.
        Add .pkl to the filename.

    Args:
        obj (object): Object to be saved.
        filename (str): Name of the pickle file.

    Examples:
        >>> from spotpython.utils.file import save_pickle
        >>> save_pickle(obj, filename="obj.pkl")
    """
    filename = filename + ".pkl"
    with open(filename, "wb") as f:
        pickle.dump(obj, f)