Add docstrings to all feedback components and add to docs
jeffkayne opened this issue · 7 comments
The feedback "collect" components are missing docstrings explaining the arguments input to each component.
These should also figure in the docs. An example of this implementation with the DataContext can be seen here:
- docstrings: https://github.com/trubrics/trubrics-sdk/blob/main/trubrics/context.py
- docs: https://trubrics.github.io/trubrics-sdk/data_context/
All docstrings in this library follow the Google Style Docstring
Hey @jeffkayne , I would like to start working on this issue. Can u guide me?
Hello @rishabsinghh, welcome:)
A good place to start would be to view one component, for example the dash component, and to understand the goal of the component. Running the code snippet in the README.md will give you a feel for what the code does too.
Then, as stated in the description, you can add a docstring to the function with information about what the function does, with descriptions for each function parameter too.
Thanks !
Thanks Jeff :).
Hey @jeffkayne! Just wanted to ask? In the Collect_feedback_component, what does Metadata exactly mean?
Metadata is an argument that is fed into the Feedback data model, allowing users to log any data / metrics directly into the feedback .json.
This can typically be things that the user can see on the app that a data scientist has built; like the model prediction, some feature values, explainability metrics etc.
Oh, got it! Thanks :)