This repository contains scripts and one jupyter notebook to execute benchmarks for the signac core application. In the current version, they are optionally compared against datreant.
Author: Carl Simon Adorf, csadorf@umich.edu
All code within this repository, except for the tempdir.py module, are distributed under the MIT License.
The full license text is in the LICENSE.txt file.
To run benchmarks, first clone this repository, then execute:
./run.shThis will generate a collection file called benchmark.txt, which contains the benchmark results.
To visualize and plot the benchmark results, execute the report.ipynb jupyter notebook.
The run_benchmark.py script is the main benchmark execution script.
Execute python run_benchmark.py --help to get help on how to execute benchmarks.
For example, to run a benchmark for signac for a dataspace with 1,000 directories in the /tmp directory, you would execute:
python run_benchmark.py signac -N 1000 --root=/tmp