Skip to content

Milestones

List view

  • No due date
    0/2 issues closed
  • Currently the private data interface doesn't work with dice-ml sklearn explainers. We can support for dice_random.py and dice_genetic.py. dice_kdtree.py needs to look at entire training data so for kdtree private data interface cannot be supported.

    No due date
    2/3 issues closed
  • The following methods need to be refactored to reduce the cyclometric complexity of these methods. (diceml) C:\Users\gaugup\Documents\GitHub\DiCE>flake8 . --count --max-complexity=10 --max-line-length=127 --statistics .\dice_ml\data_interfaces\private_data_interface.py:16:5: C901 'PrivateData.__init__' is too complex (11) .\dice_ml\data_interfaces\public_data_interface.py:17:5: C901 'PublicData.__init__' is too complex (18) .\dice_ml\explainer_interfaces\dice_genetic.py:114:5: C901 'DiceGenetic.do_KD_init' is too complex (11) .\dice_ml\explainer_interfaces\dice_genetic.py:389:5: C901 'DiceGenetic.find_counterfactuals' is too complex (13) .\dice_ml\explainer_interfaces\dice_KD.py:154:5: C901 'DiceKD.vary_valid' is too complex (14) .\dice_ml\explainer_interfaces\dice_pytorch.py:337:5: C901 'DicePyTorch.round_off_cfs' is too complex (11) .\dice_ml\explainer_interfaces\dice_pytorch.py:412:5: C901 'DicePyTorch.find_counterfactuals' is too complex (30) .\dice_ml\explainer_interfaces\dice_random.py:37:5: C901 'DiceRandom._generate_counterfactuals' is too complex (22) .\dice_ml\explainer_interfaces\dice_tensorflow1.py:516:5: C901 'DiceTensorFlow1.find_counterfactuals' is too complex (25) .\dice_ml\explainer_interfaces\dice_tensorflow2.py:414:5: C901 'DiceTensorFlow2.find_counterfactuals' is too complex (25) .\dice_ml\explainer_interfaces\explainer_base.py:229:5: C901 'ExplainerBase.feature_importance' is too complex (22) .\dice_ml\explainer_interfaces\explainer_base.py:521:5: C901 'ExplainerBase.is_cf_valid' is too complex (11) 12 C901 'PrivateData.__init__' is too complex (11) 12

    No due date
    3/3 issues closed