byoa 0.1.0a5

No project description provided

byoa

byoa is a command-line utility to Build Your Own Analytic based on metrics and imagery data following your business logic.

• Create your own processor from a template
• Use the processor in Docker containers or as a simple Python library
• Deploy processor in an analytics cluster

Getting started

Installation

Install via from PyPI

pip install byoa

Create the processor project

Once byoa is installed, create a new processor with:

byoa init

Several input values will be asked:

• name is the display name of the processor
• slug is the identifier of the processor
• repository is the name of the folder the project will be created in

And voila ! The new project is located in the repository folder. You should be able to:

cd repository

Building the processor

As a Python library

Then build the processor. As a Python library (wheel):

byoa build wheel

The wheel file is located in the dist folder. You can then install the library with:

pip install path_to_wheel_file.whl

As a Docker image

If you would rather build a Docker image:

byoa build image

Then run the image with:

byoa run image --mode

--mode determines the entrypoint used to run the image. For example, --api runs the processor as a FastAPI application.

Deploy the processor

Deploying a processor does several things :

• Registers the processor in an Analytics Cluster (if it is the first deployment)
• Registers the processor's version to deploy in an Analytics Cluster
• Builds and push the processor image in a container registry

The version is set in the VERSION file. You may have to set your Python path (PYTHONPATH environment variable) to src.

Deploy your processor with the following command:

byoa deploy

You undo the deployment with the command:

byoa delete --version VERSION

Contact

• Report a bug or request a feature: Issue
• For any additional information, please email us
• Socials : LinkedIn, X

License

Distributed under the MIT License.