release v0.10.1

• Bugfix for dashboard, should now work on remote compute
• Bugfix for raw explanation generation with feature maps - engineered data should no longer be attached
• Documentation fixes

release v0.10.0

• updated interpret-core to 0.1.21
• import ABC from collections.abc instead of collections for Python 3 compatibility
• fixed timeseries error where index column not removed for surrogate model when using reset_teacher on evaluation

patch release v0.9.2

• another fix for doc build due to missing docs on properties
• fix mimic method to mention the surrogate model type

patch release v0.9.1

• fix doc build errors

release v0.9.0

• make sure all data is returned as a list from explanation
• fix save for inner explainer and wrapped model in kernel explainer
• fix aggregation logic to use mean

patch release v0.8.1

• validate inputs in dashboard
• fix does quack for sparse local importance values

release v0.8.0

• added full e2e sparse linear surrogate model support
• localize dashboard and widget
• fix num classes sparse handling
• added sparse local importance values aggregation logic from engineered to raw features

release v0.7.0

• python doc updates
• fix notebook failures due to ports in use from previous notebooks, add automatic port scan mechanism by default
• fix typo causing explanation dashboard failure in databricks
• fix logger causing failure if pytorch not installed
• changed plotly chart table to have display block
• fix port not assigned issue
• improved gevent error message
• initialize icons once, fix exceptions if empty object passed in
• resilience to inconsistently shaped input data

release v0.6.0

• updated shap and interpret-core dependencies
• support timeseries models
• address accessibility bugs
• improve error handling in widget to propagate inner exception and stack trace
• notebook test infrastructure improvements

patch release v0.5.1

patch release v0.5.1

• fix explanation dashboard failing to load and serialization error on notebooks