Skip to main content

Numerical regression testing

Project description

nrtest

PyPI Package Build Status Documentation Status

nrtest is an end-to-end regression testing framework, designed for scientific software that perform numerical calculations.

Features

nrtest aims to simplify your workflow:

  • JSON files describe the software under test and the tests themselves

  • result files are stored in a portable benchmark directory

  • benchmarks are compared by iterating through tests and their results

  • custom comparisons can easily be added through extensions

Basic Usage

As an example usage, we consider testing TOPAS. This is a Monte Carlo tool for particle simulation, designed for medical physics research. Of course, such a tool must be rigorously validated against experimental data. But it is also useful to frequently run shorter tests, checking for regressions by comparing results to a previous version.

First, we describe the software under test in a configuration file called apps/topas-2.0.3.json. Note that setup_script defines the environment needed to run the software.

{
    "name": "topas",
    "version" : "2.0.3",
    "setup_script" : "/path/to/topas-2.0.3/setup.sh",
    "exe" : "topas"
}

We then describe the test in second configuration file called tests/Scoring_01.json. In doing so, we define the command-line arguments presented to the executable and the input files needed for the test to run. Finally, we also specify the expected output files, and declare how they should be compared to a benchmark. Here we use topas binned, which is a custom comparison routine, though some comparison routines are bundled with nrtest. It is also easy to add your own.

{
    "name": "Scoring_01",
    "version": "1.0",
    "description": "Basic test shooting a 6cm diameter proton beam into a water phantom.",
    "args": [
        "Scoring_01.txt"
    ],
    "input_files": [
        "Scoring_01.txt",
        "GlobalParameters.txt"
    ],
    "output_files": {
        "Dose.csv": "topas binned"
    }
}

To execute the test, we tell nrtest where to find the configuration files and where to output the benchmark. Note that nrtest will search tests/ for tests, though we could have specified tests/Scoring_01.json.

$ nrtest execute apps/topas-2.0.3.json tests/ -o benchmarks/2.0.3
INFO: Found 1 tests
Scoring_01: pass
INFO: Finished

To compare to a previous benchmark:

$ nrtest compare benchmarks/2.0.3 benchmarks/2.0.2
Scoring_01: pass
INFO: Finished

More advanced usage is detailed in the documentation.

History

0.2.5 (2021-08-10)

  • Add ability to skip tests based on minimum_app_version config

0.2.4 (2019-09-11)

  • Fix sporadic failures on Windows when deleting temporary files

0.2.3 (2018-02-09)

  • Add ability to write comparison results to JSON file

0.2.0 (2016-02-21)

  • Add extensions for compare functions

0.1.0 (2016-02-21)

  • First release on PyPI.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

nrtest-0.2.5.tar.gz (24.7 kB view details)

Uploaded Source

Built Distribution

nrtest-0.2.5-py2.py3-none-any.whl (15.2 kB view details)

Uploaded Python 2 Python 3

File details

Details for the file nrtest-0.2.5.tar.gz.

File metadata

  • Download URL: nrtest-0.2.5.tar.gz
  • Upload date:
  • Size: 24.7 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.4.2 importlib_metadata/4.6.3 pkginfo/1.7.1 requests/2.26.0 requests-toolbelt/0.9.1 tqdm/4.62.0 CPython/3.9.6

File hashes

Hashes for nrtest-0.2.5.tar.gz
Algorithm Hash digest
SHA256 211a0981ca54efd910637f26ed9941a5a4345e16f6865027ad224f06290deb37
MD5 2fab2ac6eeef453793d295dc42065d29
BLAKE2b-256 8daad359524e1dc6183cc1cc1c25b0425848860d37a3138592b6c4384b272a26

See more details on using hashes here.

Provenance

File details

Details for the file nrtest-0.2.5-py2.py3-none-any.whl.

File metadata

  • Download URL: nrtest-0.2.5-py2.py3-none-any.whl
  • Upload date:
  • Size: 15.2 kB
  • Tags: Python 2, Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.4.2 importlib_metadata/4.6.3 pkginfo/1.7.1 requests/2.26.0 requests-toolbelt/0.9.1 tqdm/4.62.0 CPython/3.9.6

File hashes

Hashes for nrtest-0.2.5-py2.py3-none-any.whl
Algorithm Hash digest
SHA256 cc6db77f9aa67509a14905395d3c8c97e4813a3c2b118cec184cb5e76320cf77
MD5 0c91915e1b4bf8b1f31553dcdaecb305
BLAKE2b-256 906b035b80f4dde115cb2b33f08c21427536d5fb3d5e27d37d676de417ff96e6

See more details on using hashes here.

Provenance

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page