taskchain 1.1.1

Utility for running data and ML pipelines

TaskChain

Documentation

Docs

https://flowerchecker.gitlab.io/taskchain/

Install

pip install taskchain

From source

git clone https://github.com/thran/taskchain
cd taskchain
python setup.py install
# or
python setup.py develop

Chengelog

1.1.1

• improve chain representation in jupyter
• add tasks_df parameter to chains
• add support for uses in contexts (same syntax as in configs)
• improve create_readable_filenames
  • use config name as default name
  • better verbose mode
• force method of both Chain and Task now supports delete_data parameter which delete persisted data
  • it defaults to False
  • be careful with this
• add Makefile

1.1.0

• release to PIP

1.0.3

• more types can be used for run method, e.g. dict or Dict[str, int]
• forbid some names of parameters with special meaning in configs (uses, tasks, ...)
• you should import from taskchain instead of taskchain.taks, later is deprecated and will be removed
  • use from taskchain import Task, Config, Chain or import taskchain as tc; tc.Task
• MultiChain are now more robust, you can use them with configs with context, and it will work correctly

Development

Develop docs

run server which dynamically serves docs web.

mkdocs serve

Build docs

Create documentation as static files. Used automatically by GitLab CI.

mkdocs build --strict --verbose