Better tests

celer/tests/test_dense.py

@@ -6,7 +6,7 @@
 
 
 def build_dataset(n_samples=50, n_features=200, n_informative_features=10,
-                  n_targets=1):
+                  n_targets=1, sparse_X=False):
     """
     build an ill-posed linear regression problem with many noisy features and
     comparatively few samples
@@ -17,34 +17,44 @@ def build_dataset(n_samples=50, n_features=200, n_informative_features=10,
     else:
         w = random_state.randn(n_features)
     w[n_informative_features:] = 0.0
-    X = random_state.randn(n_samples, n_features)
-    y = np.dot(X, w)
-    X_test = random_state.randn(n_samples, n_features)
-    y_test = np.dot(X_test, w)
+    if sparse_X:
+        X = sparse.random(n_samples, n_features, density=0.5, format='csc',
+                          random_state=random_state)
+        X_test = sparse.random(n_samples, n_features, density=0.5,
+                               format='csc', random_state=random_state)
+    else:
+        X = random_state.randn(n_samples, n_features)
+        X_test = random_state.randn(n_samples, n_features)
+    y = X.dot(w)
+    y_test = X_test.dot(w)
     return X, y, X_test, y_test
 
 
 def test_celer_path_dense():
-
+    """Test Lasso path computation on dense data."""
     X, y, _, _ = build_dataset(n_samples=50, n_features=50, n_targets=1)
 
     alpha_max = np.max(np.abs(X.T.dot(y)))
     n_alphas = 10
     alphas = alpha_max * np.logspace(0, -2, n_alphas)
 
     tol = 1e-6
-    betas, thetas, gaps = celer_path(X, y, alphas=alphas, tol=tol, verbose=1)
+    alphas, coefs, gaps, thetas = celer_path(X, y, alphas=alphas, tol=tol,
+                                             return_thetas=True)
+    betas = coefs.T
     np.testing.assert_array_less(gaps, tol)
 
 
 def test_celer_path_sparse():
-
-    X, y, _, _ = build_dataset(n_samples=50, n_features=50, n_targets=1)
-    X = sparse.csc_matrix(X)
+    """Test Lasso path computation on sparse data."""
+    X, y, _, _ = build_dataset(n_samples=50, n_features=50, n_targets=1,
+                               sparse_X=True)
     alpha_max = np.max(np.abs(X.T.dot(y)))
     n_alphas = 10
     alphas = alpha_max * np.logspace(0, -2, n_alphas)
 
     tol = 1e-6
-    betas, thetas, gaps = celer_path(X, y, alphas=alphas, tol=tol, verbose=1)
+    alphas, coefs, gaps, thetas = celer_path(X, y, alphas=alphas, tol=tol,
+                                             return_thetas=True)
+    betas = coefs.T
     np.testing.assert_array_less(gaps, tol)