added total_samples to return of sample_neighbors_in_scoregraph

graphmuse/samplers/__init__.py

@@ -184,17 +184,20 @@ def sample_neighbors_in_score_graph(note_array, depth, samples_per_node, targets
 	samples_per_layer: PyList(type=np.ndarray, length=depth+1)
 		List of numpy arrays of nodes (called layers) where the last layer corresponds to 'targets' and each n-th layer which isn't the last is a subset of the pre-neighborhood of the n+1-th layer
 	edges_between_layers: PyList(type=np.ndarray(2, N), length=depth)
-		List of numpy arrays of edges which show how 2 consecutive layers in samples_per_layer are connected 
+		List of numpy arrays of edges which show how 2 consecutive layers in samples_per_layer are connected
+	total_samples : numpy.ndarray
+		the union of samples_per_layer
 	"""
 	assert len(targets)>0
 
 	onsets = note_array["onset_div"].astype(numpy.int32)
 	durations = note_array["duration_div"].astype(numpy.int32)
-	samples_per_layer, edges_between_layers = c_sample_neighbors_in_score_graph(onsets, durations, depth, samples_per_node, targets)
+	samples_per_layer, edges_between_layers, total_samples = c_sample_neighbors_in_score_graph(onsets, durations, depth, samples_per_node, targets)
 	# move to torch tensors
 	samples_per_layer = [torch.from_numpy(layer) for layer in samples_per_layer]
 	edges_between_layers = [torch.from_numpy(edges) for edges in edges_between_layers]
-	return samples_per_layer, edges_between_layers
+	total_samples = torch.from_numpy(total_samples)
+	return samples_per_layer, edges_between_layers, total_samples
 
 
 def sample_preneighbors_within_region(cgraph, region, samples_per_node=10):

graphmuse/samplers/sampling_sketch.py

@@ -149,7 +149,7 @@ def sample_from_graphlist(self, graphlist):
 
             if self.sample_rightmost:
                 # Sample rightmost node-wise by num layers (because of reverse edges missing)
-                right_layers, edges_between_right_layers = csamplers.sample_neighbors_in_score_graph(random_graph.note_array, self.num_layers-2, self.samples_per_node, right_extension)
+                right_layers, edges_between_right_layers, total_right_samples = csamplers.sample_neighbors_in_score_graph(random_graph.note_array, self.num_layers-2, self.samples_per_node, right_extension)
             else:
                 right_layers, edges_between_right_layers = [], []
             edges_between_layers = torch.cat((left_edges, right_edges, edges_between_right_layers, edges_between_left_layers), dim=1)
@@ -158,8 +158,7 @@ def sample_from_graphlist(self, graphlist):
 
             # Translate edges to subgraph indices (do this on GPU when available).
             subgraph_edge_index = torch.cat((edges_within_region, edges_between_layers), dim=1).to(self.device)
-            # TODO: add total right samples
-            sampled_nodes = torch.cat((torch.arange(region[0], region[1]), sampled_nodes_first_layer, total_left_samples)).to(self.device)
+            sampled_nodes = torch.cat((torch.arange(region[0], region[1]), sampled_nodes_first_layer, total_left_samples, total_right_samples)).to(self.device)
             new_mapping = torch.arange(sampled_nodes.shape[0], device=self.device)
             nodes_remap = torch.empty_like(random_graph.x.shape[0]).to(self.device)
             nodes_remap[sampled_nodes] = new_mapping

include/musical_sampling.c

@@ -349,6 +349,10 @@ static PyObject* sample_neighbors_in_score_graph(PyObject* csamplers, PyObject*
 	HashSet node_hash_set;
 	HashSet_new(&node_hash_set, (Index)PyArray_SIZE(prev_layer));
 
+	HashSet total_samples;
+	HashSet_new(&total_samples, (Index)PyArray_SIZE(prev_layer));
+	HashSet_init(&total_samples);
+
 	HashSet node_tracker;
 	HashSet_new(&node_tracker, samples_per_node);
 
@@ -406,6 +410,7 @@ static PyObject* sample_neighbors_in_score_graph(PyObject* csamplers, PyObject*
 			if(neighbor_count <= samples_per_node){
 				for(Index j=lower_bound; j<upper_bound; j++){
 					HashSet_add_node(&node_hash_set, j);
+					HashSet_add_node(&total_samples, j);
 
 					edge_list[2*edge_list_cursor]=j;
 					edge_list[2*edge_list_cursor+1]=i;
@@ -435,6 +440,7 @@ static PyObject* sample_neighbors_in_score_graph(PyObject* csamplers, PyObject*
 					Node j = perm[rand_ix];
 
 					HashSet_add_node(&node_hash_set, j);
+					HashSet_add_node(&total_samples, j);
 
 					edge_list[2*edge_list_cursor]=j;
 					edge_list[2*edge_list_cursor+1]=i;
@@ -458,6 +464,7 @@ static PyObject* sample_neighbors_in_score_graph(PyObject* csamplers, PyObject*
 					}
 
 					HashSet_add_node(&node_hash_set, j);
+					HashSet_add_node(&total_samples, j);
 
 
 					edge_list[2*edge_list_cursor]=j;
@@ -480,11 +487,14 @@ static PyObject* sample_neighbors_in_score_graph(PyObject* csamplers, PyObject*
 		PyList_SET_ITEM(edges_between_layers, layer-1, (PyObject*)edges);
 	}
 
+	PyArrayObject* np_total_samples = HashSet_to_numpy(&total_samples);
+
+	HashSet_free(&total_samples);
 	HashSet_free(&node_tracker);
 	HashSet_free(&node_hash_set);
 	free(edge_list);
 
-	return PyTuple_Pack(2, samples_per_layer, edges_between_layers);
+	return PyTuple_Pack(3, samples_per_layer, edges_between_layers, np_total_samples);
 }
 
 static PyObject* sample_preneighbors_within_region(PyObject* csamplers, PyObject* args){