Added tests for checking if in an epoch the samples generated are cor…

…rect.

tests/mapper/test_link_mappers.py

@@ -320,6 +320,20 @@ def test_edge_consistency(shuffle):
     #             feature_size=2 * self.n_feat,
     #         )
 
+    def test_GraphSAGELinkGenerator_not_Stellargraph(self):
+        G = nx.Graph()
+        elist = [(1, 2), (2, 3), (1, 4), (3, 2)]
+        G.add_edges_from(elist)
+
+        # Add example features
+        for v in G.nodes():
+            G.node[v]["feature"] = np.ones(1)
+
+        with pytest.raises(TypeError):
+            GraphSAGELinkGenerator(
+                G, batch_size=self.batch_size, num_samples=self.num_samples
+            )
+
     def test_GraphSAGELinkGenerator_zero_samples(self):
 
         G = example_Graph_1(self.n_feat)
@@ -411,9 +425,9 @@ def test_GraphSAGELinkGenerator_isolates(self):
         ne, nl = gen[0]
         assert pytest.approx([1, 1, 2, 2, 4, 4]) == [x.shape[1] for x in ne]
 
-    def test_GraphSAGELinkGenerator_unsupervisedSampler(self):
+    def test_GraphSAGELinkGenerator_unsupervisedSampler_flow(self):
         """
-        This tests link generator's iterator for real time link generation i.e. there is no pregenerated list of samples provided to it.
+        This tests link generator's initialization for on demand link generation i.e. there is no pregenerated list of samples provided to it.
         """
         n_feat = 4
         n_batch = 2
@@ -432,12 +446,62 @@ def test_GraphSAGELinkGenerator_unsupervisedSampler(self):
             unsupervisedSamples
         )
 
-        # UnsupervisedSampler object or a list of samples is not passed
+        # The flow method is not passed UnsupervisedSampler object or a list of samples is not passed
+        with pytest.raises(TypeError):
+            gen = GraphSAGELinkGenerator(
+                G, batch_size=n_batch, num_samples=n_samples
+            ).flow("not_a_list_of_samples_or_a_sample_generator")
+
+        # The flow method is not passed nothing
         with pytest.raises(TypeError):
             gen = GraphSAGELinkGenerator(
                 G, batch_size=n_batch, num_samples=n_samples
             ).flow()
 
+    def test_GraphSAGELinkGenerator_unsupervisedSampler_sample_generation(self):
+
+        G = example_Graph_2(self.n_feat)
+
+        rw = UniformRandomWalk(G)
+
+        unsupervisedSamples = UnsupervisedSampler(G, walker=rw)
+
+        mapper = GraphSAGELinkGenerator(
+            G, batch_size=self.batch_size, num_samples=self.num_samples
+        ).flow(unsupervisedSamples)
+
+        assert mapper.data_size == 8
+        assert self.batch_size == 2
+        assert len(mapper) == 4
+
+        for batch in range(len(mapper)):
+            nf, nl = mapper[batch]
+
+            assert len(nf) == 3 * 2
+            assert len(set(mapper.head_node_types)) == 1
+
+            for ii in range(2):
+                assert nf[ii].shape == (
+                    min(self.batch_size, mapper.data_size),
+                    1,
+                    self.n_feat,
+                )
+                assert nf[ii + 2].shape == (
+                    min(self.batch_size, mapper.data_size),
+                    2,
+                    self.n_feat,
+                )
+                assert nf[ii + 2 * 2].shape == (
+                    min(self.batch_size, mapper.data_size),
+                    2 * 2,
+                    self.n_feat,
+                )
+                assert len(nl) == min(self.batch_size, mapper.data_size)
+                assert sorted(nl) == [0, 1]
+
+        with pytest.raises(IndexError):
+            nf, nl = mapper[4]
+
 
 class Test_HinSAGELinkGenerator(object):
     """