skore 0.1.0

Tooling and assistance for data scientists to "Own Your Data Science"

skore

Installation

For now, the only supported method to use skore is from source. Follow the instructions in CONTRIBUTING.md to install dependencies and start the UI.

Quick start

For a complete introductory example, see our basic usage notebook. The resulting skore report has been exported to this HTML file: download it and open it in your browser to visualize it.

In your shell, run the following to create a project file project.skore (the default) in your current working directory:

python -m skore create

Run the following in your Python code to load the project:

from skore import load

project = load("project.skore")

You can save items you need to track in your project:

project.put("my int", 3)

You can also get them back:

project.get("my int")

By default, strings are assumed to be Markdown:

project.put("my string", "Hello world!")

Note that put overwrites previous data

project.put("my string", "Hello again!")

You can list all the keys in a project with:

project.list_keys()

You can delete items with:

project.delete_item("my int")

Then, in the directory containing your project, run the following command to start the UI locally:

python -m skore launch project.skore

This will automatically open a browser at the UI's location.

Roadmap

With Skore, you can:

• Store data
• Visualize data

In the future, you will be able to:

• Share visualizations of your data
• Extract insights from your data
• Get tips on how to improve your data science code

Contributing

See CONTRIBUTING.md for more information and to contribute to the evolution of this library.