sweepexp 1.2.0

A python package for running parallel experiments across parameter grids with MPI.

Features

• Flexible Experimentation with Parameter Grids: SweepExp simplifies running experiments over a grid of parameter combinations. Results are stored in an xarray dataset for easy access and analysis.
• Parallelization: Support for parallelization using multiprocessing, or MPI for high-performance computing.
• Combining Sweeps: Run multiple different sweeps together with a shared pool of workers using SweepGroup.
• User-Friendly API: Define the function to be tested, set up parameter sweeps, and specify return types effortlessly.

Installation

SweepExp can be installed via pip:

pip install sweepexp

Usage

The followin example shows how to setup a simple experiment that is run on a grid of parameters. Where each parameter combination is run in parallel on separate processes.

from sweepexp import SweepExpParallel

# Define a function to be tested
def my_custom_experiment(x: float, y: float) -> dict:
    """Add and multiply two numbers."""
    return {"addition": x + y, "multiplication": x * y}

sweep = SweepExpParallel(
    func = my_custom_experiment,
    parameters = { "x": [1, 2], "y": [3, 4, 5] },
)

sweep.run()

print(sweep.data)

with the output:

<xarray.Dataset> Size: 160B
Dimensions:         (x: 2, y: 3)
Coordinates:
    * x               (x) int64 16B 1 2
    * y               (y) int64 24B 3 4 5
Data variables:
    status          (x, y) <U1 24B 'C' 'C' 'C' 'C' 'C' 'C'
    addition        (x, y) float64 48B 4.0 5.0 6.0 5.0 6.0 7.0
    multiplication  (x, y) float64 48B 3.0 4.0 5.0 6.0 8.0 10.0

For more information on how to use the package, please refer to the documentation

How to cite

@software{Rosenau_sweepexp_2025,
          author = {Rosenau, Silvano Gordian},
          doi = {10.5281/zenodo.14779187},
          month = jan,
          title = {{SweepExp: A python package for running parallel experiments across parameter grids.}},
          url = {https://github.com/Gordi42/sweepexp},
          version = {1.0.2},
          year = {2025}
}

Author

• Silvano Gordian Rosenau

License

SweepExp is licensed under the MIT License. See LICENSE for more information.