Data Quality Dashboard Optimisation

• Data Quality Dashboard: add dataset overview widget
• Data Quality Dashboard: add correlations widget
• Speeded uploading via preview plots optimisation
• Paging in Data Quality feature table widget

DataQuality Dashboard and Profile

• DataQualityTab() is now available for Dashboards! The Tab helps to observe data columns, explore their properties and compare datasets.
• DataQualityProfileSection() is available for Profiles as well.
• ColumnMapping update: added task parameter to specify the type of machine learning problem. This parameter is used by DataQualityAnalyzer and some data quality widgets to calculate metrics and visuals with the respect to the task.

Support for strings as a categorical features

• Added monitors for NumTargetDrift, CatTargetDrift, ClassificationPerformance, ProbClassificationPerformance
• Fixed RegressionPerformance monitor
• Supported strings as a categorical features in DataDrift and CatTargetDrift dashboards
• Supported boolean features in DataDrift dashboard

QualityMetricsOptions

Analyzers Refactoring: analyzer result became a structured object instead of a dictionary for all Analyzers

The following Quality Metrics Options are added:

• conf_interval_n_sigmas (the width of confidence intervals ): int = DEFAULT_CONF_INTERVAL_SIZE
• classification_treshold (the threshold for true labels): float = DEFAULT_CLASSIFICATION_TRESHOLD
• cut_quantile (cut the data by right, left and two-sided quantiles): Union[None, Tuple[str, float], Dict[str, Tuple[str, float]]] = None

Backward compatibility for Dashboard and Profile

Added backward compatibility for imports:

• Widgets and Tabs can be imported from evidently directly, but this is deprecated behavior and cause warning
• Sections can be imported from evidently directly, but this is deprecated behavior and cause warning

Repository Restructuring

• Library source code is moved to the src/evidently folder
• Docs, Tests, and Examples are moved to the top level of the repo
• Widgets and Tabs are moved inside of the src/evidently/dashboard folder, as those are parts of the Dashboard
• Sections are moved inside of the src/evidently/model_profile folder, as those are parts of the Model_profiles
• Docs are stored in the repo: docs/book folder
• DataDriftAnalyzer refactoring: analyzer results became a structured object instead of a dictionary

No Input Data Modification

• fixed: input DataFrames cannot be changed during any calculations (fixed by making shallow copies)
• fixed: chi-square statistical test uses normalized frequencies (with respect to the latest scipy version)
• current dataset is optional for Performace Tabs and Sections calculation (None value can be passed)
• improved readme

Custom Stattest & Confidence level for features

Data Drift Options:

• Created confidence: Union[float, Dict[str, float]] - option can take a float or a dict as an argument. If float has passed, then this confidence level will be used for all features. If dict has passed, then specified features will have a custom confidence levels (all the rest will have default confidence level = 0.95)
• Updated nbinsx: Union[int, Dict[str, int]] - option can take an int or a dict as an argument. If int has passed, then this number of bins will be used for all features. If dict has passed, then specified features will have a custom number of bins (all the rest will have default number of bins = 10)
• Updated feature_stattest_func: Union[None, Callable, Dict[str, Callable]] - option can take a function or a dict as an argument. If a function has passed, then this function will be used to measure drift for all features. If dict has passed, then for specified features custom functions will be used (all the rest features will be processed by an internal algorithm of drift measurement)

Package building:

• Fixed dependencies

Сustom Widgets and Tabs

• Support widgets order for include_widgets parameter
• Support an ability to add a custom widget to Tabs with include_widgets parameter
• Moved options to a separate module
• Added options to specify statistical tests for DataDrift and TargetDrift Dashboards:
  stattest_func - to set a custom statistical test for all the features
  feature_stattest_func - to set a custom statistical tests for each individual feature
  cat_target_stattest_func - to set a custom statistical test for categorical target
  num_target_stattest_func - to set a custom statistical test for numerical target
• Refactored Widgets and Tabs for simpler customisation

Customizable Dashboards

• Supported custom list of Widgets for Tabs in Dashboard with help of verbose_level and include_widgets parameters
• Added parameter verbose_level: 0 - to create a Tab with the shortest list of Widgets, 1 - to create a full Tab
• Added parameter include_widgets : ["Widget Name 1", "Widget Name 2", etc]. This parameter overwrites verbose_level (if both are specified) and allows to set a custom list of Widgets
• Added Tab.list_widgets() method to list all the available Widgets for the current Tab
• Created Options entity to specify Widgets and Tabs customisable settings
• Created ColumnMapping entity instead column_mapping python dictionary