svm.py

from benchopt import BaseSolver

from sklearn.svm import SVC


# The benchmark solvers must be named `Solver` and
# inherit from `BaseSolver` for `benchopt` to work properly.
class Solver(BaseSolver):

    # Name to select the solver in the CLI and to display the results.
    name = 'SVM'

    # List of parameters for the solver. The benchmark will consider
    # the cross product for each key in the dictionary.
    # All parameters 'p' defined here are available as 'self.p'
    # and are set to one value of the list.
    parameters = {
        'kernel': ['linear', 'poly', 'sigmoid'],
    }

    # List of packages needed to run the solver. See the corresponding
    # section in objective.py. Here `scikit-learn` is already present
    # so no need to add it again.
    requirements = []

    def set_objective(self, X_train, y_train):
        # Define the information received by each solver from the objective.
        # The arguments of this function are the results of the
        # `Objective.get_objective`. This defines the benchmark's API for
        # passing the objective to the solver.
        # It is customizable for each benchmark.
        self.X_train, self.y_train = X_train, y_train
        self.clf = SVC(kernel=self.kernel)

    def run(self, _):
        """Run the solver.

        Parameters
        ----------
        _ : ignored
            With sampling_strategy="run_once", this parameter is unused.
        """
        # This is the method that is called to fit the model.
        self.clf.fit(self.X_train, self.y_train)

    def get_result(self):
        # Returns the model after fitting.
        # The output of this function is a dictionary whose keys define the
        # keyword arguments for `Objective.evaluate_result`.
        # This defines the benchmark's API for solvers' results.
        # It is customizable for each benchmark.
        return dict(model=self.clf)