def test_vector_space(): X, y = get_data() evo = EvoMSA(evodag_args=dict(popsize=10, early_stopping_rounds=10, n_estimators=3), models=[['EvoMSA.model.Corpus', 'EvoMSA.model.Bernulli']]) evo.model(X) nrows = len(X) X = evo.vector_space(X) assert X[0].shape[0] == nrows
def test_EvoMSA_kfold_decision_function(): from sklearn.preprocessing import LabelEncoder X, y = get_data() le = LabelEncoder().fit(y) y = le.transform(y) evo = EvoMSA(evodag_args=dict(popsize=10, early_stopping_rounds=10, n_estimators=3), models=[['EvoMSA.model.Corpus', 'EvoMSA.model.Bernulli']]) evo.model(X) X = evo.vector_space(X) cl = evo.models[1][1] D = evo.kfold_decision_function(cl, X[1], y) assert len(D[0]) == 4 assert isinstance(D[0], list)
def test_EvoMSA_fit_svm(): from sklearn.preprocessing import LabelEncoder X, y = get_data() from sklearn.svm import LinearSVC from EvoMSA.model import Bernoulli model = EvoMSA(stacked_method_args=dict(popsize=10, early_stopping_rounds=10, n_estimators=3), models=[['EvoMSA.model.Corpus', 'EvoMSA.model.Bernoulli']], n_jobs=2) le = LabelEncoder().fit(y) y = le.transform(y) model.model(X) Xvs = model.vector_space(X) model.fit_svm(Xvs, y) print(model._svc_models) assert len(model._svc_models) == 2 for ins, klass in zip(model._svc_models, [LinearSVC, Bernoulli]): assert isinstance(ins, klass)