Fixes #447: forward sklearn estimator attributes (#574)

Author: charavelg

tslearn/svm/svm.py

@@ -13,6 +13,27 @@
 
 
 class TimeSeriesSVMMixin:
+
+    @property
+    def support_(self):
+        check_is_fitted(self, ['svm_estimator_', '_X_fit'])
+        return getattr(self, "svm_estimator_").support_
+
+    @property
+    def dual_coef_(self):
+        check_is_fitted(self, ['svm_estimator_', '_X_fit'])
+        return getattr(self, "svm_estimator_").dual_coef_
+
+    @property
+    def coef_(self):
+        check_is_fitted(self, ['svm_estimator_', '_X_fit'])
+        return getattr(self, "svm_estimator_").coef_
+
+    @property
+    def intercept_(self):
+        check_is_fitted(self, ['svm_estimator_', '_X_fit'])
+        return getattr(self, "svm_estimator_").intercept_
+
     def _preprocess_sklearn(self, X, y=None, fit_time=False):
         force_all_finite = self.kernel not in VARIABLE_LENGTH_METRICS
         if y is None:
@@ -446,9 +467,6 @@ class TimeSeriesSVR(TimeSeriesSVMMixin, RegressorMixin,
     intercept_ : array, shape = [1]
         Constants in decision function.
 
-    sample_weight : array-like, shape = [n_samples]
-        Individual weights for each sample
-
     svm_estimator_ : sklearn.svm.SVR
         The underlying sklearn estimator
 

tslearn/tests/test_svm.py

@@ -1,11 +1,16 @@
 import numpy as np
 
+import pytest
+
+from sklearn.exceptions import NotFittedError
+
 from tslearn.metrics import cdist_gak
 from tslearn.svm import TimeSeriesSVC, TimeSeriesSVR
 
 __author__ = 'Romain Tavenard romain.tavenard[at]univ-rennes2.fr'
 
 
+
 def test_gamma_value_svm():
     n, sz, d = 5, 10, 3
     rng = np.random.RandomState(0)
@@ -22,3 +27,22 @@ def test_gamma_value_svm():
         cdist_mat = cdist_gak(time_series, sigma=np.sqrt(gamma / 2.))
 
         np.testing.assert_allclose(sklearn_X, cdist_mat)
+
+def test_attributes():
+    n, sz, d = 5, 10, 3
+    rng = np.random.RandomState(0)
+    time_series = rng.randn(n, sz, d)
+    labels = rng.randint(low=0, high=2, size=n)
+
+    for ModelClass in [TimeSeriesSVC, TimeSeriesSVR]:
+        linear_model = ModelClass(kernel="linear")
+
+        for attr in ['coef_', 'support_', 'support_vectors_',
+                     'dual_coef_', 'coef_', 'intercept_']:
+            with pytest.raises(NotFittedError):
+                getattr(linear_model, attr)
+
+        linear_model.fit(time_series, labels)
+        for attr in ['coef_', 'support_', 'support_vectors_',
+                     'dual_coef_', 'coef_', 'intercept_']:
+            assert hasattr(linear_model, attr)