* ckpt
* refactor + unit tests
* auto lint fixes
* ckpt
* od label UI ckpt
* od agg label working ckpt
* working backend
* UI update to remove pred label
* added UI tests
* lint fixes
* lint fixes
* MO test tweak for vision
* comment fixes
* comment fixes
* auto lint fixes
* reverted MO test changes
* auto lint fixes
* Add postga build trigger (#1755) (#1756)
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
* Add model wrapper for wrapping predictions and test data (#1762)
* Add model wrapper for wrapping predictions and test data
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
* Fix failing tests
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
* Change description of cohort selection panel in Aggregate Feature Importance (#1770)
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
* Support cohort filtering of string target in rai_insights (#1771)
* Port tests
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
* Fix rai_insights
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
* Simplify tests in test_cohort_filter.py (#1772)
* Simply tests in test_cohort_filter.py
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
* Change test name
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
* Forecasting Dashboard
This commit adds the Forecasting Dashboard capabilities in both UI and SDK.
* remove unrelated file
* basic set of changes to get it working again
* forecasting_grains -> time_series_id_column_names
* fix breaking changes in latest main, add quantile prediction, fix date conversion, remove obsolete managers
* fix UI components (including table and what-if creation, chart still left tbd)
* get dashboard to work with multiple time series and switching back and forth
* remove data explorer and model overview, create and edit cohort functionalities from forecasting
* lintfix
* localizzation
* remove console.log
* remove question
* remove unrelated changes, fix package.json
* string fixes and package.json adjustment
* localization
* localization
* more localization and removing of unrelated files
* remove unrelated files
* remove unrelated file changes
* remove is_forecasting_true_y
* lintfix
* add dropdown label
* feature metadata input validation updates
* cache transformation predictions
* lintfix
* remove isUndefinedOrEmpty
* remove isUndefinedOrEmpty
* rename isAllDataCohort file as requested
* lintfix
* first draft of notebook
* get notebook to run, generate cohorts (not finished)
* ntoebook
* convert filter args and column names properly
* fix datetime handling
* fix forecast request handling
* working version for bug bash
* add wheels
* add UI changes
* update notebook to fix model download
* 15 iterations
* lintfix
* add example of operation method string
* add check for context
* add explanatory comment
* move description creation logic into getDescription
* lintfix
* remove datetime_features from snapshot
* lintfix and other weirdnesses
* undo print removal
* remove wheels, lintfix, small cleanup
* use META_JSON directly instead of declaring a variable
* move constants to common files and reuse
* lowercase f
* isort
* remove datetime_features
* sktime notebook (1st half)
* isort
* debugging the notebook issue
* use parentheses instead of backslashes
* flake8
* isort
* sktime progress
* pre-unified RAIInsights commit
* pre-unified RAIInsights commit 2
* first pass over unification
* unify RAIInsights and RAIForecastingInsights
* adjustments to serialization
* cleanup, test fixes, consolidation, TODO fixes
* fix tests
* isort
* flake8
* feature metadata and categorical_features consolidation
* fix serialization of feature metadata
* fix tests
* flake8
* tests for feature metadata and cat features
* flake8
* remove remnants of RAIForecastingInsights
* fix string
* move other utils back to rai insights
* undo ModelTask import changes
* undo ModelTask references to raiutils
* undo references to ModelTask in raituils
* remove raiutils changes since they're in a separate branch
* datetime_features and time_series_id_features replacement
* pytest.mark.skip
* furthe datetime_features fixes
* TS ID col name fix
* try out asfreq
* notebook try asfreq interpolation
* attempt to condense index into a single col
* add hidden feature flag, fix tests
* remove experimental file, isort
* undo string change
* restructure wrappers and field to file mapping code
* remove unnecessary TODO
* get AML notebook to work
* get AML notebook fully working
* resolve some issues with sktime
* notebook
* remove duplicated test case
* flake8
* isort
* flalke8
* wrap zip object with list
* use ModelTask from raiutils
* remove prints
* flake8
* isort
* use FORECASTING from raiutils.models
* isort
* fix sktime issue
* get past Python errors
* fix predictions array shape
* fix UI-Python mismatch with renamed time and time series ID column names (now called "features")
* lintfix
* undo changes that aren't needed any longer
* remove AML notebook
* update notebook by removing unnecessary content
* notebook tests, basic setup
* ids for tests
* fix notebook based on latest changes in the repo
* add test for adding a what if case
* flake8 and cleanup
* isort
* lintfix
* fix test
* move notebook
* fix nb path
* fix kernelspec
* add pmdarima dependency
* skip forecasting notebook on py 3.6
* remove commented out line
* skip explanation
* add reason kw arg name
* lintfix
* fix case for vision, text
* fix syntax error
* fix python version comparison since it doesn't work with <=
---------
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
Co-authored-by: Gaurav Gupta <47334368+gaugup@users.noreply.github.com>
Co-authored-by: Gaurav Gupta <gaugup@microsoft.com>
Co-authored-by: John Wang <t-wangjohn@microsoft.com>
* fix optional method validation
* strftime
* fix optional method handling
* error handling in model outputs method
* small change to avoid try except
* lintfix
* add tests for optional/required methods being present or not and succeeding/failing
* constant for time format
* ea changes
* added unit test
* OD model & data ckpt
* test fix
* torch dependency
* removed 3.6 from PR gate
* removed 3.6 from PR gate
* test fix
* import fixes
* str format fix
* test update
* revert unneeded changes
* test ckpt
* image utils
* reverted test
* test reverts
* reverted rai test utils
* test fix
* lint fixes
* enum fixes
* reverting rai insight
* added classes & model task
* reverted change
* Change `ValueError` to `UserConfigValidationException`
The `ValueError` being raised is asking user to update the configuration. So this should be `UserConfigValidationException`.
* Fix failing test and linting
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
---------
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
* add validation for the classification scenario when a user input a float-type target column
* use is_float_dtype
* add check if the float can be converted to integer
* Port over logic of `load_adult_income_dataset` from `dice-ml` to unblock gates
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
* Fix linting and add TODO
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
* Add fixture for dataset fetch
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
---------
Signed-off-by: Gaurav Gupta <gaugup@microsoft.com>
* Raise value error for min max computation of feature range
* Update rai_insights.py
* Add test
* Update test name
* Update formatting
* Update exception type
* Update test for new exception type
* Update formatting
Ilya Matiach
Gaurav Gupta
Advitya Gemawat
Roman Lutz
hawestra
tongy-msft
kicha0