You can install the latest version by running pip install --upgrade cobalt-ai

Features:

Group comparison in the UI now supports a choice of different statistical tests for numerical features.
In addition to the t-test, the Kolmogorov-Smirnov test and the Wilcoxon rank-sum test are supported, as well as a version of the t-test that uses permutation sampling to approximate the p-value instead of the t-distribution.
Workspace.get_group_neighbors() is a new method that finds a group af nearby neighbors of a given CobaltDataSubset. This neighborhood group can also be used as Group B in the group comparison UI.
The graph layout algorithm has been substantially improved and now presents cleaner, easier-to-read graphs. Some configuration options are available in cobalt.settings.
- Adjacent nodes are now prevented from overlapping. This makes the relationships between strongly connected nodes much easier to see. If needed, this can be disabled by changing the value of cobalt.settings.graph_prevent_node_overlaps.
- Connected components of the graph are now laid out in order of size from left to right. Smaller components are allocated less space, and components are packed to make efficient use of space.
- Graphs are now rotated to best fit the available space.
- The interactive graph animation is now run in a separate thread and does not interfere with other UI updates.

Quality of Life Improvements:

Embeddings for text data can now be created directly by calling CobaltDataset.add_text_column_embedding().
Prompts given by register_license() are now clearer and more informative.
GroupResultsCollection is now a subclass of GroupCollection, making it easier to explore and analyze failure groups and clusters.

Fixes:

A bug that could cause timestamps to display incorrectly when used in graph coloring was fixed.

Examples visualizing before and after v0.3.9:

New Release: Cobalt Version 0.3.9

Next Generation AI Model Evaluation