Evidently Community Call #2 Recap: custom text comments, color schemes and a library of statistical tests

May 24, 2022
ama with alexey grigorev
In April, we held the second call with the Evidently Community to give a tour of the recent feature updates in our open-source ML monitoring tool. As Evidently itself is built for community and with the community, connecting with users is our primary inspiration source. Thanks to everyone for joining!

And in case you missed it, below is a recap of what we covered.

Do you want to join the next community calls? Sign up here to get notified about the next one.

More customization features: text comments and color schemes

We introduced the new customization options available in Evidently 0.1.47 and above:

Custom color schemes

With this feature, you can add some color to the Evidently Dashboards. You can change the default red and gray colors used for the two datasets and customize the palette for all the smaller details that appear in other widgets. The feature is handy if you want the dashboards to suit your corporate color guidelines or just feel creative!

Here is an example and a pro tip by Alex Strick van Linschoten: you can use tools like Color Hunt to find a matching combination.
Custom text comments

Now you can also create text annotations in Dashboards, and include them like a custom widget. Add notes, leave comments or describe report results to share thoughts and discuss issues with other members of your team.

More statistical tests to detect and target data drift

We added more options on how to detect target drift and data drift. Previously, you needed to implement a custom function to replace the default statistical tests. Now, you can directly choose between some popular statistical tests from the options library. These include Wasserstein Distance test, Jensen–Shannon distance, Kullback-Leibler divergence, and Population Stability Index (PSI).

Use this example Jupyter notebook to check out the new features or dive into the color schema and statistical tests docs for more details.

To follow the Q&A session and watch the full replay of Evidently Community Call #2, hit play on the recording:

How can I contribute?

We welcome all sorts of contributions, not only the code one!

Give feedback. Recent personalization options are directly inspired by the user feedback, so do not hesitate to send us feature requests. Post to our Discord community or create an issue on GitHub.

Contribute to the codebase. We are building Evidently together with the community! If you want to take part, check out the contribution guide, and feel free to hop on our Discord to chat about your contribution ideas. You can also open an issue on GitHub to report a bug or suggest improvements.

And if you like Evidently, spread the word!

Want to join the next Evidently Community Call?

Sign up for our Community Call Newsletter and stay updated on future community calls.

Join us for the Community Call #3 and the live coding session to learn about the recent drift detection algorithm update. Register here to get a friendly reminder and an invite straight to your calendar. We are looking forward to seeing you all at Evidently Community Call #3!
At Evidently AI, we create open-source tools to analyze and monitor machine learning models. Check our project on GitHub and give it a star if you like it!

Want to stay in the loop?
Community Manager

You Might Also Like: