feat: add regression branch

GNNs/PyG/hgat_node_classification.ipynb

@@ -404,7 +404,7 @@
     "            out_units = out_dim if i == (num_layers - 1) else hidden_dim\n",
     "            heads = 1 if i == (num_layers - 1) else num_heads\n",
     "            self.layers.append(\n",
-    "                GATConv(in_units, out_units, heads=heads, dropout=dropout)\n",
+    "                GATConv(in_units, out_units, heads=heads, dropout=dropout, add_self_loops=False)\n",
     "            )\n",
     "\n",
     "    def reset_parameters(self):\n",
@@ -709,7 +709,7 @@
     "            out_units = out_dim if i == (num_layers - 1) else hidden_dim\n",
     "            heads = 1 if i == (num_layers - 1) else num_heads\n",
     "            self.layers.append(\n",
-    "                GATConv(in_units, out_units, heads=heads, dropout=dropout)\n",
+    "                GATConv(in_units, out_units, heads=heads, dropout=dropout, add_self_loops=False)\n",
     "            )\n",
     "\n",
     "    def reset_parameters(self):\n",

GNNs/Spektral/gcn_node_classification.ipynb

@@ -389,7 +389,7 @@
     "optimizer = Adam(learning_rate=hp[\"lr\"])\n",
     "loss_fn = CategoricalCrossentropy()\n",
     "\n",
-    "one_hot_y = to_categorical(y)\n",
+    "one_hot_y = to_categorical(y, num_classes=7)\n",
     "tf_a = tf.SparseTensor(#converts the scipy sparse matrix to a tensorflow sparse matrix\n",
     "    indices=np.array([adj.row, adj.col]).T,\n",
     "    values=adj.data,\n",
@@ -673,13 +673,13 @@
    "outputs": [],
    "source": [
     "def preprocess_batch(graph):\n",
-    "  x, adj, y, mask_tr, mask_va, mask_te = graph.x, graph.A, graph.y, graph.train_mask, graph.val_mask, graph.test_mask\n",
-    "  one_hot_y = to_categorical(y)\n",
-    "  tf_a = tf.SparseTensor(#converts the scipy sparse matrix to a tensorflow sparse matrix\n",
-    "    indices=np.array([adj.row, adj.col]).T,\n",
-    "    values=adj.data,\n",
-    "    dense_shape=adj.shape)\n",
-    "  return x, tf_a, one_hot_y\n",
+    "    x, adj, y, mask_tr, mask_va, mask_te = graph.x, graph.A, graph.y, graph.train_mask, graph.val_mask, graph.test_mask\n",
+    "    one_hot_y = to_categorical(y, num_classes=7)\n",
+    "    tf_a = tf.SparseTensor(#converts the scipy sparse matrix to a tensorflow sparse matrix\n",
+    "        indices=np.array([adj.row, adj.col]).T,\n",
+    "        values=adj.data,\n",
+    "        dense_shape=adj.shape)\n",
+    "    return x, tf_a, one_hot_y\n",
     "\n",
     "val_acc_metric = tf.keras.metrics.CategoricalAccuracy()\n",
     "val_loss_metric = tf.keras.metrics.CategoricalCrossentropy()"
@@ -918,7 +918,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.9.13"
+   "version": "3.9.16"
   },
   "vscode": {
    "interpreter": {

README.md

@@ -79,6 +79,7 @@ We recommend starting with the tutorials in the `basics` folders if you are new
 | basics | [pyTigergraph_101.ipynb](./basics/pyTigergraph_101.ipynb)  | Basic pyTigerGraph examples|
 | basics | [gsql_101.ipynb](./basics/gsql_101.ipynb)  | Basic GSQL 101 using pyTigerGraph |
 | basics | [gsql_102.ipynb](./basics/gsql_102.ipynb)  | Advanced GSQL 102 (pattern match) using pyTigerGraph |
+| basics | [template_query.ipynb](./basics/template_query.ipynb) | How to call template query with pyTigerGraph |
 
 ### 2. Graph Algorithms