microsoft
/
responsible-ai-toolbox
зеркало из https://github.com/microsoft/responsible-ai-toolbox.git
1
0
Форкнуть 0
Код Задачи Пакеты Проекты Релизы Вики Активность
responsible-ai-toolbox/CHANGES.md

152 KiB
Исходник Ответственный История

Changes

This file tracks changes to the raiwidgets package. It should be updated in PRs that make changes, not separately. At the top of the file will always be a v-next entry to add changes that were made since the last release. Each set of changes should be grouped by

  • educational materials
  • new features
  • breaking changes
  • bug fixes and tests
  • other

Note that it is not required to have an entry for every pull request. Instead, please try to add only changes that are meaningful to users who read this file to understand what changed.

v-next

  • educational materials
  • new features
  • breaking changes
  • bug fixes and tests
  • other

v0.30.0

v0.29.0

v0.28.0

v0.27.0

v0.26.0

v0.25.0

v0.24.1

v0.24.0

v0.23.0

v0.22.0

v0.21.0

v0.20.0

v0.19.0

v0.18.2

v0.18.1

v0.18.0

v0.17.0

  • educational materials
    • add e2e test steps to CONTRIBUTING.md file for developers/contributors
    • add steps for validating UX and SDK changes to documentation
    • replace boston housing dataset with california housing dataset in tests and example notebook due to ethical concerns and biases
    • update dnn pytorch notebook example to give more interesting results by increasing model epochs
  • new features

    • Responsible AI Dashboard

      • move performance metric chart, disaggregated metric table, and StatisticsUtils into core-ui
      • add highchart, all components will be moved to highchart instead of plotly gradually
      • add metrics to StatisticsUtils - F1 score, selection rate, mean absolute error
      • add cohort and filter definitions in raiwidgets SDK
      • enable highchart exporting functionality

    • Responsible AI Dashboard for Text

      • add new interpret-text library skeleton

    • Causal

      • add treatment bar chart
  • bug fixes and tests

    • Responsible AI Dashboard

      • add e2e UI tests for Causal Analysis - classification tests
      • add e2e UI tests for Feature Importance - aggregate importances
      • add e2e UI tests for Feature Importance - individual importances
      • add e2e UI tests for What-If Counterfactuals
      • add e2e UI tests for Data Explorer - diabetes-regression-model-debugging notebook
      • add e2e UI tests for Causal Analysis - diabetes-regression-model-debugging notebook
      • add e2e UI tests for Model Overview - diabetes-regression-model-debugging notebook
      • add e2e UI tests for responsibleaidashboard-housing-classification-model-debugging notebook
      • fix e2e UI What-if sort test
      • add e2e UI tests for Feature Importance - responsibleaidashboard-diabetes-decision-making notebook
      • add e2e UI tests for What-if, Model Overview and Data explorer - responsibleaidashboard-diabetes-decision-making notebook
      • add e2e UI tests for Aggregate Causal Analysis - diabetes-decision-making notebook
      • add multiclass classification dataset to UI test app and set up basic model-assessment tests for Model Overview
      • add notebook test for pytorch example notebook and add pytorch to CI
      • move to new fluentUI dependency
      • correctly annotate python warnings
      • separate out raiwidgets and UI tests into its own workflow, and added new pipeline to run RAIwidget e2e tests
      • add js and raiwidget code coverage
      • add upload step to e2e test for CI RAIWidgets Python Typescript pipeline
      • add upload step to e2e test to CI notebooks pipeline
      • return exception details to UI
      • add docstrings for locale parameter

    • RAIInsights

      • catch dirty data conditions in RAI Insights validation
      • update responsibleai to interpret-community 0.24.0
      • update responsibleai to interpret-community 0.24.1
      • update responsibleai to interpret-community 0.24.2
      • make responsibleai managers public and add back compat connectors
      • make explanations save behavior similar to other managers
      • change logic to walk categorical columns in validate_train_test_categories
      • cover missing test scenario in state_directory_management.py
      • default categorical feature to array
      • add error analysis test for regression task in RAIInsights
      • make causal manager add and compute behavior similar to other managers

    • Counterfactual

      • add feature importance description for counterfactual local importance
      • update Counterfactual explanation text description

    • Error Analysis

      • fix importances taking too long to calculate for large data in error analysis
      • add a perf test for validating training the surrogate error tree on large data and a minor optimization for traversing the tree
      • add filtering on index to error analysis backend for RAI dashboard cohort switch
      • remove unnecessary multiple calls to tree api
      • refactor tree view state to be private static in tree view renderer and not in parent
      • create github action to release erroranalysis
      • add excludes cohort filter test in erroranalysis
      • refactor matrix filter and area state to be private static
      • make search case insensitive in feature list
      • add more docs for error analysis surrogate tree
      • optimize tree traversal logic in error analysis to reduce execution time when running on several million rows

    • Interpret

      • allow ExplanationDashboard to show limited view for more than 1k features instead of erroring

    • Causal

      • improve documentation for nuisance_model and heterogeneity_model
      • implement list method in causal manager

  • other

    • fix pypi releases by reverting back to pypi token per package
    • fix dummy counterfactual feature importance values for UI test app
    • add pseudo language for localization test
    • integrate codecov into responsible-ai-toolbox repository
    • remove test warning due to use of deprecated np.object
    • add python 3.9 tests
    • fix ipython dependency in rai-core-flask and remove from raiwidgets
    • update setup.py files to carry python 3.9 classifiers
    • fix broken builds due to pip upgrade which broke pip-tools
    • remove mac temporarily from rai widgets step to fix failing builds
    • remove node-version from CI-python.yml
    • fix flaky flask test
    • improve description for responsibleai and raiwidgets pypi packages

v0.16.0

  • educational materials
    • major readme refactoring changes + filename changes for public release of RAI Dashboard
    • add missing doc for pred_y_dataset and min_child_samples
    • add npm and yarn installation instructions
  • new features

    • Counterfactual

      • counterfactual list style updates: highlight counterfactual examples columns with different color, and bold size
      • highlight first and last row of counterfactual list
      • implement list method for counterfactual manager

    • Error Analysis

      • add scrollbar area around matrix cells when there are many categories
      • add warning message for downsampled explanation data in ErrorAnalysis
  • bug fixes and tests

    • Responsible AI Dashboard

      • add e2e UI tests for Dataset Explorer - whisker plot
      • add e2e UI tests for Dataset Explorer - scatter plot
      • add e2e UI tests for Dataset Explorer - flyout
      • add e2e UI tests for Dataset Explorer - cohort scenarios
      • add e2e UI tests for Model Statistics - chart tests
      • refactor binding this with arrow func
      • add error message for duplicate cohort name
      • handle context cohort delete cases for data explorer

    • RAIInsights

      • throw error on mutating predict function in RAIInsights
      • use spilt orientation to save train and test dataframe to address duplicate indices condition in them
      • add .json extension to dtypes files and add tests
      • add predict and predict_proba outputs to RAIInsights save method
      • refactor RAIInsights load function into smaller functions
      • refactor classifier detection logic into common function
      • fix issue with missing directories
      • gracefully handle failure in loading model.pkl from disk
      • remove unused test dependencies responsibleai
      • add missing unit tests for _convert_to_list()
      • separate out save() and load() tests
      • add more unit tests for serialize_json_safe()
      • update responsibleai to interpret-community 0.23.0

    • Counterfactual

      • add e2e UI tests for what-if counterfactuals - creating counterfactuals

    • Error Analysis

      • fix coverage calculation for cell tooltips in heatmap for recall, precision and f1 score metrics
      • fix handling numpy types in heatmap for categories

    • Interpret

      • improve state management for explainer manager
      • test for save/load/save bug in Explanations
  • other
    • fix top broken image link in tour notebook on github
    • add isort to build pipelines
    • fix SubBarChart flaky test
    • update notebook tests for new notebooks
    • add flake8-bugbear for better python code contructs
    • add flake8-breakpoint to avoid code checkin with active breakpoints
    • remove csv files from source for cognitive services example notebooks
    • add yarn e2e-watch shortcut
    • replace existing user token with shared account token for releases

v0.15.1

  • educational materials
    • Notebooks updated for new class names (ResponsibleAIDashboard etc.)
  • new features
    • None
  • breaking changes

    • RAI Insights/Model Analysis

      • Rename explainers directory to generators when saving RAI Insights object
  • bug fixes

    • RAI Insights/Model Analysis

      • Fix Issue #1046, related to a save/load/save cycle when no explainer present
  • other
    • Extra tests for WhatIf counterfactuals
    • Checks on the UUID directory structure of RAI Insights

v0.15.0

  • new features

    • Model Analysis

      • move raw data into /data directory
      • perform schema validations during compute and load
      • rename model analysis to rai insights
      • rename ModelAnalysisDashboard to ResponsibleAIDashboard
      • directory based state management for model analysis tools
      • add .json extension to train and test dataset

    • Error Analysis

      • port ErrorReport schema validation for error analysis to responsibleai
      • add error analysis error importances (aka correlation of features with error) to model analyzer API
      • add support for special classification outcome filter to error analysis backend to work with model analysis filters

    • Counterfactual

      • remove save as new data point option for read-only mode on What-if counterfactual
      • split counterfactual output into individual json files
      • add support for non-string categorical values
  • bug fixes

    • Model Analysis

      • fix cohort text out of button box
      • fix bug when the tree node has 0 errors and N correct, the raw data table shows the opposite with N+1 errors and -1 correct.
      • when classes are not passed, sort them automatically in RAIInsights
      • fix composite filter issues

    • Counterfactual

      • fix sort feature columns by counterfactual

    • Error Analysis

      • fix map shift text not being clear, provide conditional dialog text based on current view
      • fix tree map connection line doubled after switching to dark theme

  • other

    • update readme links to be absolute instead of relative path
    • add more UX IFI tests for responsibleaitoolbox-classification-model-assessment notebook
    • several E2E UX test fixes
    • add E2E UX tests for Feature Importance - Cohort functionality
    • add test to render ModelAnalysisDashboard after save and load
    • add test to render ResponsibleAIDashboard after save and load
    • add E2E tests for What-if Counterfactuals - common functionalities & Y Axis Flyout
    • add separate github workflow for python linting

v0.14.0

  • educational materials
    • add Cognitive Service fairness assessment notebooks
  • new features

    • Error Analysis

      • add new accuracy and f1-score metrics in erroranalysis python backend and UI frontend
      • add importances to ErrorReport for error analysis
      • make ErrorReport, PredictionsAnalyzer and ModelAnalyzer public in erroranalysis package
      • display message bar after saving a cohort in error analysis dashboard

    • Counterfactual

      • implement save and load for counterfactual manager
      • make load method uniform across all RAI managers
      • perform schema validations on serialized counterfactual output

    • Interpret

      • upgrade interpret-community to 0.22.0
  • bug fixes

    • Model Analysis

      • remove drag and drop instructions in the dashboard navigation
      • add theme support for model statistics and dataset explorer
      • remove setup_explainer after ModelAnalysis load call in tests
      • add theme support for feature importances, causal, counterfactual tabs
      • hide opened feature list when clicking on the heatmap tab in model analysis
      • add localization for delete cohort dialog

    • Interpret

      • set is_run class variable in explainer manager

    • Causal

      • add init file to causal schema directory

    • Error Analysis

      • fix color for metrics that are not error metrics to be green in legend and heatmap nan cells
      • update features list in static view to have apply button removed and labels disabled for tree view parameters
      • update features list text for static case
      • fix running what if on categorical numeric values in error analysis dashboard
      • add theme support for error analysis view
      • fix for importances in static view causing failures when unspecified
      • fix feature list alignment issue

  • other

    • add documentation for the release process
    • add E2E test for model analysis, running from python jupyter notebook and validating in cypress
    • sort all python imports using isort
    • add aggregate feature importance tests for responsibleaitoolbox
    • add E2E individual feature importance tests for responsibleaitoolbox-classification-model-assessment notebook
    • fix image links in responsibleaitoolbox-classification-model-assessment.ipynb notebook to reference images in folder instead of using embedded images directly
    • rename causal tests

v0.13.0

  • new features

    • Model Analysis

      • add "Save amd switch" button to cohort creation
      • pass classes parameter to both error analysis dashboard and error analysis manager in model analysis
      • upgrade interpret-community to 0.20.0

    • Error Analysis

      • update erroranalysis to 0.1.24
  • bug fixes

    • Model Analysis

      • fix issue with widgets rendering with 8px height

    • Error Analysis

      • fix precision and recall score failing on string labels in error analysis
      • fix multiclass precision and recall calculation for NA zero values in heatmap
      • edit string in error analysis cohort creation
      • remove text in heatmap suggesting to select features when in static view
      • minor doc fix for examples in error analysis dashboard

    • Counterfactual

      • fix horizontal scroll bar disappears for small screen in what-if panel

    • Fairness

      • fix data issue caused robot page exception on fairnessDetails
  • other
    • fix broken link to getting started notebook
    • fix broken image link on main readme
    • fix the issues with sample notebooks markdown images
    • delete old notebooks
    • change tests to run on new notebooks, add new notebook to tests

v0.12.1

  • new features

    • Error Analysis

      • add disabled tree view and heatmap for static view
  • bug fixes

    • Error Analysis

      • fix class length causing labels to not be shown in simple view of Error Analysis
      • fix cohort filter tooltip behind cohort info panel

    • Causal

      • fix trace0 appearing on hover tooltip in plots

    • Counterfactual

      • handle cases where counterfactuals failed to compute
      • add tooltip for counterfactual info
  • other
    • some grammatical errors were corrected on documentation page

v0.12.0

  • new features

    • Model Analysis

      • add global cohort statistics

    • Error Analysis

      • add metric selector dropdown
      • add tree view parameters as sliders to features selector panel
      • add tooltip to heatmap categories and left-justify the heatmap

    • Counterfactual

      • upgrade dice-ml to 0.7.2
  • bug fixes

    • Model Analysis

      • fix cohort information in heatmap resizing in model analysis
      • log warning when regression model has predict_proba()

    • Error Analysis

      • fix error rate gradient for classification scenario
      • fix quantile binning toggle should stay on when making actions
      • fix multiple issues in instance view, including:
        • when selecting multiple rows in correct and incorrect panels, and going to all selected, if a point is removed in the middle all of the points below get removed
        • when removing points in the all selected panel, the count does not get updates in the correct/incorrect panels until they are selected
        • when using what if, make points not selectable
      • fix precision and recall calculation in the UI changing with number of bins
      • decrease the lower bound on first category for quantile binning
      • fix handling of numeric features names in pandas for error analysis tree view
      • reverse y axis in heatmap so that higher values are towards the top of the matrix
      • add warnings for 2d heatmap case and test for quantile binning with duplicate edges

    • Causal

      • sort causal data point by effect size
      • set treatment list spin range
      • fix causal global effect plot x-axis titles cut off
      • fix causal categorical treatment average gains text

    • Counterfactual

      • change current class label for regression task
      • raise exception when users configures less than required counterfactuals
      • show options for categorical features on counterfactual list

    • Fairness

      • add error handling for missing metrics in fairness
  • other
    • remove Axios library from dependencies
    • rename wrapped-flask to rai_core_flask
    • remove deprecated babelconfig option in workspace.json

v0.11.0

  • new features
    • table for list of available cohorts
    • added more metrics to ErrorAnalysisDashboard, including precision, recall and mean abs error
  • bug fixes
    • fix various UI usability bugs
    • fix for clicking custom scatter plot point

    • Error Analysis

      • fix selecting what if datapoint crashing dashboard
      • fix error in quantile binning when there are too many duplicate points
      • fix displayed error rate in all data cohort in cohort list
      • fix error in heatmap for classification case
      • fix features list not getting disabled in static view
      • update description text for static view
  • other
    • improve validation for model prediction
    • reorder counterfactual table columns
    • add absolute sorting for local importance charts
    • improve spacing and fonts across dashboard
    • fix typos in docstrings
    • improve wording of text on dashboard
    • add warnings for soft data limits
    • remove font override for all components
    • add code coverage to repository

v0.10.0

  • educational materials

    • update current model analysis classification notebook by adding treatment features for causal analysis

  • new features

  • breaking changes

  • bug fixes

    • Model Analysis:

      • change icon for ViewList in main menu
      • fix model statistics component not showing all cohorts
      • put causal component last, remove navigation
      • hide removal dialog on click
      • move cohort info and button to sticky menu, replace error stats with filters

    • Error Analysis:

      • refactor heatmap code and fix display issues for categoricals
      • fix incorrect values in heatmap when there are empty error cells
      • added quantile binning and number of bins functionality to erroranalysis package
      • added quantile binning and number of bins to UX and hooked up calls to backend
      • added tooltip to list of features for tree view
      • fix backend to support equals filter for integer categorical columns
      • fix backend to support include and exclude filters for integer categorical columns
      • update default value for matrix feature
      • fix cohort when switching tree view

    • Causal:

      • remove unnecessary z-index on what-if and causal components
      • update causal policy to support categorical features
      • format policy tree categorical features
      • fix description on how confounding features are handled
      • use lighter lines for causal plotly point graphs

    • Fairness:

      • incorporate uncertainty quantification in fairness widget v2

    • Counterfactuals:

      • upgrade dice-ml to 0.7
      • update counterfactual list to allow string values
      • extend unseen categorical check to counterfactuals
    • add dropdown to dataset explorer by default
    • fix removing what if point will keep a faded start on the chart
    • suppress distracting logging from flask when running widgets
    • fix large x and y axis going out of bounds
    • refactor multiple instances of convert to list method in several dashboards
    • add missing interpretability information and guidelines

  • other

    • fix message and check for environment in release pipeline

v0.9.4

  • bug fixes
    • fixed causal treatment policy display issue happening when switching between features

v0.9.3

  • educational materials
  • new features
    • causal treatment cost argument can be passed per feature/sample/category
    • target added to counterfactual tabular output
    • causal analysis now trains the model on training data, not both train and test
    • min_child_samples parameter for error analysis
  • breaking changes
  • bug fixes
    • improved text for causal what-if
    • fix for missing values in dataset
    • improvement for default explanation tab
    • fix for multiple feature importances detail view
    • fix for causal what-if slider being stuck
    • fixes for cohort selection
    • fixes for error analysis matrix view
    • fix for extra scroll bar in counterfactuals view
    • fix for train data input in classification notebook
    • fix for bug in counterfactual feature_to_vary parameter
    • fix for x and y axes of error analysis heatmap
    • fix for counterfactual predict
  • other
    • improved error messages for responsibleai ModelAnalysis
    • improved validation when not passing a model to ModelAnalysis
    • updated ModelAnalysis classification notebook to use separate train and test datasets

v0.9.2

  • educational materials
  • new features
    • Model Analysis:
      • add search box to counterfactuals panel
      • add support for feature to vary and permitted range in counterfactual manager
    • add matrix features and tree features to error report object in erroranalysis package
  • breaking changes
  • bug fixes
    • call flask service credential
    • Model Analysis:
      • fix hardcoded feature name in counterfactual panel
      • fix refresh on global cohort change
      • add tree_features and matrix_features params to static tree and heatmap
      • update sticky menu
      • change counter factual desired_range to tuple
    • fix aggregate feature importance chart for multiclass classification
    • fix serialize_json_safe not escaping special characters which caused dashboard to fail to render on pandas dataframe that had double quotes in string values
  • other
    • add dice-ml to intersphinx_mapping
    • add missing documentation for counterfactual manager
    • fix python responsibleai package test for windows

v0.9.1

  • educational materials
  • new features
  • breaking changes
    • error report in erroranalysis package had json_tree renamed to tree and json_matrix renamed to matrix
  • bug fixes
    • fix econML inputs
    • add categorical feature to causal whatif
    • fix counterfactual panel scroll bar
    • round precision in causal analysis tables
    • fix calculating feature importances after error report due to dataframe on analyzer being modified
  • other
    • dump information about the most time consuming tests

v0.9.0

  • educational materials
    • add Model Analysis notebook for regression
  • new features
    • local importance chart for counterfactual dashboard
    • add multi causal policy support
    • add delete tab dialog and remove inline widget for model assessment dashboard
    • add treatment policy and table to causal analysis dashboard
    • create new custom individual feature importance component
  • breaking changes
  • bug fixes
    • Model Analysis:
      • sticky layer for cohort and dashboard settings
      • restrict EA tree to 80% width
      • restrict ipykernel<0.6
      • fix flyout title and description
      • fix add button behavior by adding components in the appropriate spot
      • show add button only for components that can render
      • add validations for input parameters for Model Analysis class
      • validate treatment features in causal manager
      • explicitly pass model task to MimicExplainer in model analyzer
    • Error Analysis:
      • remove 100k row limit
      • fix zero nodes in tree view
      • fix search functionality in features list
      • fix displayed filter order in tree view to start from root node
      • raise warning for older version of pandas when feature names contain '-'
      • expand number of decimals shown in filter tooltips when values below 0.1
      • fix regression case for 1d heatmap matrix metrics being calculated incorrectly
    • set original data from dropdown for counterfactuals
    • use test data instead of train data for error analysis manager in model analyzer
    • causal analysis to keep all levels in dataframe outputs
    • pin networkx pip package to prevent installation of matplotlib
    • add target_column causal policy gain to UI control
  • other
    • unit tests for multiclass classification for causal analysis
    • add model analysis notebooks to notebook gate
    • add test for causal manager for None categoricals
    • fix example code for model assessment

v0.8.0

  • educational materials
  • new features
    • enabled categoricals on MimicExplainer for the explainer manager
    • updated causal constants to enable linear models and allow high cardinality features by default
    • support newest policy output format from econml
  • breaking changes
    • continuous_features in CounterfactualManager.add() is deprecated in favor of categorical_features in ModelAnalysis
  • bug fixes
    • fixed error analysis add after model analysis deserialization
  • other
    • improved tests for visualization dependencies

v0.7.0

  • educational materials
  • new features
    • added widget serialization for causal analysis
  • breaking changes
  • bug fixes
  • other

v0.6.1

  • educational materials
  • new features
  • breaking changes
  • bug fixes
  • other

v0.6.0

  • educational materials
  • new features
  • breaking changes
  • bug fixes
  • other

v0.5.0

  • educational materials
    • cleanup commented out explainers from interpretability notebook
  • new features
    • added support for causal, counterfactual, and error analysis tools in responsibleai
    • added model analysis widget to raiwidgets
    • added support for regression to error analysis
    • added policy tree to causal manager
  • breaking changes
    • renamed package raitools to responsibleai
    • renamed top-level class RAIAnalyzer to ModelAnalysis
    • removed support for fairness in responsibleai package
  • bug fixes
    • fixed bug in precision statistic calculation in dashboard
    • fixed bug in loading explanations with ModelAnalysis
    • fixed on predict method to use correct dtypes
    • fixed individual importances chart when no data available
  • other
    • create combined feature importances tab for local and global explanation in model assessment
    • add section headers to model assessment dashboard
    • add manual cohort creation to the model assessment dashboard

v0.4.0

  • educational materials
    • cleanup commented out explainers from interpretability notebook
  • new features
    • fix databricks environment with raiwidgets dashboards
    • updates to ErrorAnalysisDashboard:
      • improve the features list to use DetailsList
      • added tooltips to legend
      • initial static view for error analysis, specifically matrix filter and tree view
      • fixed tooltip node hover
      • fix indexing issue on cohort shift
      • add new simple dashboard implementation, which only takes in predictions and does not need model or dataset
    • updates to ModelAssessmentDashboard:
      • set up tabs on the left side
      • add model statistics tab
      • add pivot for error analysis
      • remove fullscreen button
    • refactor error analysis python code into separate common package
  • breaking changes
  • bug fixes
    • reduce the lower-bound on the lightgbm dependency
    • update greenlet and gevent dependencies to fix releases
  • other
    • use global context in interpret, error-analysis, and model-assessment
    • moved dataset explorer into a new project dataset-explorer
    • created new projects causality and counterfactuals
    • remove fairness v1 dashboard

v0.3.1

  • Bug fix:
    • fix public VM failing requests in rai-core-flask package, update raiwidgets to latest rai-core-flask package

v0.3.0

  • Bug fixes:
    • Fix all data cohort always appearing first in cohorts lists
    • Add number selected text under local explanation radio buttons
    • Fix the issue that categorical feature is not able to change in cohort editor.
    • Fix categorical feature in what-if tab

v0.2.2

  • Bug fix:
    • Fix categorical feature in ice plot
    • Left align tabs in explanation window of error analysis
    • Move the selected features and feature importances inside a scrollable pane
    • Show aggregate plots by default in dataset explorer
    • Fix indexing issue in heatmap when selecting all cells

v0.2.1

  • educational materials:
    • add imports for MimicExplainer and PFIExplainer in the interpretability-dashboard-employee-attrition notebook
  • features
    • migrate cohort and dataset utilities to core-ui
    • remove circular dependencies
    • add confirmation popup in ExplanationDashboard when cancelling editing a cohort.
    • fix clipped dropdown in FairnessDashboard
    • keep FairnessDashboard chart state in single model view
    • updates to ErrorAnalysisDashboard:
      • in large data view, change cohort state to use full data instead of downsampled explanation data
      • add true y and predicted y values
      • fix categorical labels handling in what if panel
      • add breadcrumb to error explorer view and minor style adjustments to navigation panel
      • fix indexing issues for all data cohort
      • add color column to local explanation view

v0.2.0

This is the first release of the raiwidgets package.

  • educational materials:
    • notebook analyze-categoricals-binary-classification-local.ipynb added
    • notebook erroranalysis-pretability-dashboard-breast-cancer.ipynb added
    • notebook erroranalysis-interpretability-dashboard-census.ipynb added
    • notebook fairness-credit.ipynb added
    • notebook fairness-dashboard-loan-allocation.ipynb added
    • notebook fairness-interpretability-dashboard-loan-allocation.ipynb added
    • notebook interpretability-dashboard-employee-attrition.ipynb added
  • features
    • FairnessDashboard changes compared to fairlearn's FairlearnDashboard which was migrated to raiwidgets:
      • Restructured performance metric selection as list rather than set of tiles for easier navigation.
      • Added fairness metric selection using a grouped list.
      • Added dropdowns to model comparison view and single model view to allow users to quickly switch between sensitive features, performance metrics, and fairness metrics.
      • Added metric table to single model view.
      • Added dropdown to switch between charts in single model view.
      • Simplified API to extract sensitive feature names directly from sensitive_features.
    • ErrorAnalysisDashboard: first release
    • migrated ExplanationDashboard from interpret-community to raiwidgets

We're skipping v0.1.0 to get all our packages onto the same version:

  • pypi:

    • raiwidgets: first real release

  • npm:

    • @responsible-ai/fairness: previously 0.1.*
    • @responsible-ai/interpret: previously 0.0.*
    • @responsible-ai/error-analysis: first real release
    • @responsible-ai/mlchartlib: previously 0.0.*
    • @responsible-ai/core-ui: previously 0.0.*
    • @responsible-ai/localization: previously 0.0.*