Skip to content

Releases: devopshq/FuzzyClassificator

Release 1.3.72 announcement

10 May 13:25
@Tim55667757 Tim55667757
1.3.72
594a64f
Compare
Choose a tag to compare
Release 1.3.72 announcement Latest
Latest

Release version of FuzzyClassificator [1.3 build 72] uploaded to PyPI:
https://pypi.python.org/pypi/FuzzyClassificator/1.3.72

Please report new bugs or the required functionality in new tasks.

Features

  1. Now you can install FuzzyClassificator via pip:
    pip install FuzzyClassificator [--upgrade] [--pre]

  2. API work mode added. You can use Main() function if FuzzyClassificator imported as a module. Also API work in Anaconda python interpreter. Work examples with API added to README.md (See chapter "Work with API").

  3. Documentation auto-updated from README.md on the GitHub-pages. More usage examples and classificating scheme added: https://devopshq.github.io/FuzzyClassificator/

  4. --version console key added which show version as major.minor.["dev"]build. Also version is printed in the log.

  5. --show-expected console key added which show column with expected results if classifying ethalon data file.

Improvements

  1. Build-process realized using Travis-CI. After success build FuzzyClassificator auto-uploaded to PyPI: .whl archive suported.

  2. Badges with links and actual statuses from Travis-CI, PyPI and Codacy services added to README.md.

  3. A lot of code refactoring. Code quality grade increased from B to A using Codacy service: https://www.codacy.com/app/tim55667757/FuzzyClassificator/dashboard

  4. Some unit-tests on base routines added.

Assets 2
Loading

release-1.2

22 Mar 10:08
@Tim55667757 Tim55667757
release-1.2
4a95310
Compare
Choose a tag to compare
release-1.2

Important:

  1. Re-license to MIT.
  2. Freezed development in old personal project https://github.com/Tim55667757/FuzzyClassificator and move it to the Open DevOps Community: https://github.com/devopshq/FuzzyClassificator

Features:

  1. Stop train if errors less than some value.
  2. Auto-update error status in log after given epochs time.
  3. Saving the best network during training time.
  4. Updated usage examples.
  5. Optimized all critical methods - the program began to work several times faster!

See more (articles in russian):

  1. Vulnerability testing methods: https://habrahabr.ru/company/pt/blog/187636/
  2. Mathematical apparatus:
  1. Implementation: https://habrahabr.ru/company/pt/blog/274241/
  2. Op!DevOps! meet up: https://habrahabr.ru/company/pt/blog/323436/
Loading