Utility for running data and ML pipelines
Project description
TaskChain
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 supportsdelete_data
parameter which delete persisted data- it defaults to
False
- be careful with this
- it defaults to
- add Makefile
1.1.0
- release to PIP
1.0.3
- more types can be used for
run
method, e.g.dict
orDict[str, int]
- forbid some names of parameters with special meaning in configs (
uses
,tasks
, ...) - you should import from
taskchain
instead oftaskchain.taks
, later is deprecated and will be removed- use
from taskchain import Task, Config, Chain
orimport taskchain as tc; tc.Task
- use
- 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
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
taskchain-1.1.1.tar.gz
(24.7 kB
view hashes)