proceed 0.0.3

Execute pipelines based on files and containers.

Proceed

Declarative file processing with YAML and containers.

Proceed is a Python library and CLI tool for declarative batch processing. It reads a pipeline specification declared in YAML. A pipeline contains a list of steps that are based on Docker images and containers.

Each pipeline execution accepts values for declared args, allowing controlled, explicit configuration of steps at runtime. Each execution produces an execution record that accounts for accepted arg values, step logs, and checksums of input and output files.

Hopefully, Proceed will allow you to express everything you need to know about your processing pipeline in a "nothing up my sleeves" way. The pipeline specification should be complete enough to share with others who have Proceed and Docker installed. The execution record should allow for auditing of expected outcomes and reproducibility.

docs

Here are the main docs for Proceed.

Installation

Proceed requires Python and Docker to be installed. With those, it should be able to run a wide variety pipelines and steps via containers.

pip

Proceed itself is not yet available on PyPI. When it is, this will be the recommended way to install Proceed:

$ pip install proceed # TODO

git and pip

You can also install Proceed from source.

$ pip install git+https://https://github.com/benjamin-heasly/proceed.git

# editable mode
$ git checkout https://github.com/benjamin-heasly/proceed.git
$ pip install -e ./proceed

check installation

You can check if Proceed installed correctly using the proceed command.

$ proceed --version
Proceed x.y.z

$ proceed --help
usage etc...

development and testing

You can set up a development environment with conda and dev-environment.yml.

conda env create -f dev-environment.yml
# or
conda env update -f dev-environment.yml

With that, you should be able to run through the Proceed unit and integration tests.

conda activate proceed-dev
hatch run test:cov