robustx.lib package

Subpackages

• robustx.lib.distance_functions package
  • Submodules
  • robustx.lib.distance_functions.DistanceFunctions module
    • euclidean()
    • manhattan()
  • Module contents
• robustx.lib.intabs package
  • Submodules
  • robustx.lib.intabs.IntervalAbstractionPyTorch module
    • IntervalAbstractionPytorch
      • IntervalAbstractionPytorch.evaluate()
      • IntervalAbstractionPytorch.create_weights_and_bias_dictionary()
  • robustx.lib.intabs.WeightBiasDictionary module
    • create_weights_and_bias_dictionary()
  • Module contents
• robustx.lib.models package
  • Subpackages
    • robustx.lib.models.keras_models package
      • Submodules
      • robustx.lib.models.keras_models.SimpleKerasNNModel module
      • Module contents
    • robustx.lib.models.pytorch_models package
      • Submodules
      • robustx.lib.models.pytorch_models.CustomPyTorchModel module
      • robustx.lib.models.pytorch_models.SimpleNNModel module
      • Module contents
    • robustx.lib.models.sklearn_models package
      • Submodules
      • robustx.lib.models.sklearn_models.DecisionTreeModel module
      • robustx.lib.models.sklearn_models.LogisticRegressionModel module
      • robustx.lib.models.sklearn_models.SVMModel module
      • robustx.lib.models.sklearn_models.SklearnModel module
      • Module contents
  • Submodules
  • robustx.lib.models.BaseModel module
    • BaseModel
      • BaseModel._model
      • BaseModel.train()
      • BaseModel.predict()
      • BaseModel.predict_single()
      • BaseModel.predict_proba()
      • BaseModel.predict_proba_tensor()
      • BaseModel.Properties()
      • BaseModel.evaluate()
      • BaseModel.model
      • BaseModel.predict()
      • BaseModel.predict_proba()
      • BaseModel.predict_proba_tensor()
      • BaseModel.predict_single()
      • BaseModel.train()
  • robustx.lib.models.Models module
    • get_sklearn_model()
  • Module contents
• robustx.lib.tasks package
  • Submodules
  • robustx.lib.tasks.ClassificationTask module
    • ClassificationTask
      • ClassificationTask.model
      • ClassificationTask._training_data
      • ClassificationTask.get_negative_instances()
      • ClassificationTask.get_random_positive_instance()
  • robustx.lib.tasks.Task module
    • Task
      • Task._training_data
      • Task.__model
      • Task.get_negative_instances()
      • Task.get_random_positive_instance()
      • Task.model
      • Task.training_data
  • Module contents

Submodules

robustx.lib.DefaultBenchmark module

robustx.lib.DefaultBenchmark.default_benchmark(ct, methods, evaluations, subset=None, **params)

Generates and prints a table summarizing the performance of different counterfactual explanation generation methods.

@param ct: ClassificationTask. @param methods: A list or a set of method names. @param evaluations: A list or a set of evaluator names. @param subset: optional DataFrame, subset of instances you would like to generate CEs on @param **params: Additional parameters to be passed to the CE generation methods and evaluators. @return: None

robustx.lib.OptSolver module

class robustx.lib.OptSolver.OptSolver(ct)

Bases: object

A solver class that uses Gurobi to optimize a model based on a given task and instance.

Attributes / Properties

task: Task

The task to be optimized.

gurobiModel: Model

The Gurobi optimization model.

inputNodes: dict

Dictionary to store Gurobi variables for input nodes.

outputNode: Gurobi variable

The Gurobi variable representing the output node.

setup(instance, desired_output=1, delta=0.5, bias_delta=0, M=1000000000, epsilon=0.0001, fix_inputs=True):

Sets up the Gurobi model with constraints based on the provided instance and parameters.

-------

setup(instance, delta=0, bias_delta=0, M=1000, fix_inputs=True)

Sets up the Gurobi model with constraints based on the provided instance and parameters.

@param instance: pd.DataFrame or list, The input data instance for which to set up the model. @param desired_output: int, Optional, The desired output value (default is 1). @param delta: float, Optional, The delta value used in constraints (default is 0.5). @param bias_delta: float, Optional, The bias delta value used in constraints (default is 0). @param M: float, Optional, A large constant used in constraints (default is 1000000000). @param epsilon: float, Optional, The epsilon value used in constraints (default is 0.0001). @param fix_inputs: bool, Optional, Whether to fix input values or use variable bounds (default is True).

@return: None

robustx.lib.PyTorchConversions module

Module contents