simutils 0.0.2

Tools for logging simulation information for later reproductions.

simutils

Note that this package is not yet implemented.

Usage

It logs the argument of the function and other relevant information to a JSON file.

>>> from simutils import log_function

>>> @log_function
... def double_val(x):
...    return x * 2
...
>>> double_val(4)
8

Then the content of the JSON file will be:

{
    "simulation_name": "double_val_20210216_141001",
    "status": "success",
    "computer_name": "MyComputer",
    "python_version": "3.8.6 (tags/v3.8.6:db45529, Sep 23 2020, 15:52:53) [MSC v.1927 64 bit (AMD64)]",
    "cwd": "current/working/directory",
    "output_directory": "logs/double_val_20210216_141001",
    "summary_filepath": "path/to/this/file",
    "called_function": "double_val",
    "start_time": "2021-02-16 14:10:02",
    "end_time": "2021-02-16 14:10:03",
    "simulation_time": "0:00:01.012345",
    "args": {
        "x": 4
    }
}

You can pass arguments to the decorator, and you can also decorate functions in a Python script.

@log_function(
    git_repo=os.path.join(os.path.dirname(__file__), '..', '..'), # git repository to log
)
def triple_val(x):
    return x * 3

To-Do's

• Send a notification when a function execution is finished.
• Integrate with a native Python logger.
• Reproduce the original environment