Skip to main content

Experiment collection

Project description

Experiment collection

GitHub PyPI version Release

A set of utilities for storing and organizing experiments

Install

pip install experiment-collection

You can find another versions at releases or pypi.

Usage_example

Experiment - structure with experiment data;

ExperimentCollectionLocal - local experiment storage;

ExperimentCollectionRemote - remote experiment storage; could combine data from multiply sources.

from experiment_collection.experiments import Experiment, ExperimentCollectionLocal, ExperimentCollectionRemote

exps1 = ExperimentCollectionLocal('exp_test.db')
exps2 = ExperimentCollectionRemote('http://127.0.0.1/', 'exp_test')

for i in range(10):
    exp = Experiment('name_{}'.format(i))
    exp.set_metrics({'lr': 0.1})
    exp.set_params({'auc': 0.5})

    exps1.add_experiment(exp)
    exps2.add_experiment(exp)

    assert exps1.check_experiment(exp)
    assert exps2.check_experiment(exp)

# Delete latest experiment
exps1.delete_experiment(exp)
exps2.delete_experiment(exp)

assert not exps1.check_experiment(exp)
assert not exps2.check_experiment(exp)

View all results. All metrics and params auto flatten into single columns.

exps1.get_experiments()
     name                       time  params_auc  metrics_lr
0  name_0 2020-09-30 23:56:39.932871         0.5         0.1
1  name_1 2020-09-30 23:56:40.216424         0.5         0.1
2  name_2 2020-09-30 23:56:40.620029         0.5         0.1
3  name_3 2020-09-30 23:56:40.916781         0.5         0.1
4  name_4 2020-09-30 23:56:41.240535         0.5         0.1
5  name_5 2020-09-30 23:56:41.567865         0.5         0.1
6  name_6 2020-09-30 23:56:41.861890         0.5         0.1
7  name_7 2020-09-30 23:56:42.177155         0.5         0.1
8  name_8 2020-09-30 23:56:42.507883         0.5         0.1
9  name_9 2020-09-30 23:56:42.818714         0.5         0.1

License

MIT License

Copyright (c) 2020 AsciiShell (Aleksey Podchezertsev)

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

experiment_collection-0.1.5.tar.gz (6.7 kB view details)

Uploaded Source

Built Distribution

experiment_collection-0.1.5-py3-none-any.whl (8.9 kB view details)

Uploaded Python 3

File details

Details for the file experiment_collection-0.1.5.tar.gz.

File metadata

  • Download URL: experiment_collection-0.1.5.tar.gz
  • Upload date:
  • Size: 6.7 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.2.0 pkginfo/1.5.0.1 requests/2.24.0 setuptools/50.3.0 requests-toolbelt/0.9.1 tqdm/4.50.0 CPython/3.8.6

File hashes

Hashes for experiment_collection-0.1.5.tar.gz
Algorithm Hash digest
SHA256 571a0153e0878fec863d67c1a642a78df7256a185801b41d9074b93087fc54a9
MD5 dabb4ede40d91acb42e24f79b989df61
BLAKE2b-256 ad27ae779d182167f06fbf6e3574c8f08e690981cff30ac4f9d3332876d2d2f6

See more details on using hashes here.

Provenance

File details

Details for the file experiment_collection-0.1.5-py3-none-any.whl.

File metadata

  • Download URL: experiment_collection-0.1.5-py3-none-any.whl
  • Upload date:
  • Size: 8.9 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.2.0 pkginfo/1.5.0.1 requests/2.24.0 setuptools/50.3.0 requests-toolbelt/0.9.1 tqdm/4.50.0 CPython/3.8.6

File hashes

Hashes for experiment_collection-0.1.5-py3-none-any.whl
Algorithm Hash digest
SHA256 a9858b63e7dde0e8f555e24d55a91063cf54c2141581a3ba94d5d10b347552d7
MD5 1c0ada5c66368afcdf5822585935b516
BLAKE2b-256 1977cefa39048bf2228e289841da8a525ebda72ac004976ebc1f7d977167e8b0

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