[Example] Fix numpy errors in examples. (#6554)

Co-authored-by: Ubuntu <ubuntu@ip-172-31-0-133.us-west-2.compute.internal>

examples/mxnet/scenegraph/train_freq_prior.py

@@ -89,7 +89,7 @@ for _, item in train_data.items():
             for col, row in zip(cols, rows):
                 bg_matrix[gt_classes[col], gt_classes[row]] += 1
         else:
-            all_possib = np.ones_like(iou_mat, dtype=np.bool)
+            all_possib = np.ones_like(iou_mat, dtype=np.bool_)
             np.fill_diagonal(all_possib, 0)
             cols, rows = np.where(all_possib)
             for col, row in zip(cols, rows):

examples/pytorch/bgrl/eval_function.py

@@ -11,7 +11,7 @@ def fit_logistic_regression(X, y, data_random_seed=1, repeat=1):
     # transform targets to one-hot vector
     one_hot_encoder = OneHotEncoder(categories="auto", sparse=False)
 
-    y = one_hot_encoder.fit_transform(y.reshape(-1, 1)).astype(np.bool)
+    y = one_hot_encoder.fit_transform(y.reshape(-1, 1)).astype(np.bool_)
 
     # normalize x
     X = normalize(X, norm="l2")
@@ -42,7 +42,7 @@ def fit_logistic_regression(X, y, data_random_seed=1, repeat=1):
         y_pred = clf.predict_proba(X_test)
         y_pred = np.argmax(y_pred, axis=1)
         y_pred = one_hot_encoder.transform(y_pred.reshape(-1, 1)).astype(
-            np.bool
+            np.bool_
         )
 
         test_acc = metrics.accuracy_score(y_test, y_pred)
@@ -55,7 +55,7 @@ def fit_logistic_regression_preset_splits(
 ):
     # transform targets to one-hot vector
     one_hot_encoder = OneHotEncoder(categories="auto", sparse=False)
-    y = one_hot_encoder.fit_transform(y.reshape(-1, 1)).astype(np.bool)
+    y = one_hot_encoder.fit_transform(y.reshape(-1, 1)).astype(np.bool_)
 
     # normalize x
     X = normalize(X, norm="l2")
@@ -84,7 +84,7 @@ def fit_logistic_regression_preset_splits(
             y_pred = clf.predict_proba(X_val)
             y_pred = np.argmax(y_pred, axis=1)
             y_pred = one_hot_encoder.transform(y_pred.reshape(-1, 1)).astype(
-                np.bool
+                np.bool_
             )
             val_acc = metrics.accuracy_score(y_val, y_pred)
             if val_acc > best_acc:
@@ -93,7 +93,7 @@ def fit_logistic_regression_preset_splits(
                 y_pred = np.argmax(y_pred, axis=1)
                 y_pred = one_hot_encoder.transform(
                     y_pred.reshape(-1, 1)
-                ).astype(np.bool)
+                ).astype(np.bool_)
                 best_test_acc = metrics.accuracy_score(y_test, y_pred)
 
         accuracies.append(best_test_acc)

examples/pytorch/dimenet/qm9.py

@@ -181,7 +181,7 @@ class QM9(QM9Dataset):
             dist = np.linalg.norm(R[:, None, :] - R[None, :, :], axis=-1)
             # keep all edges that don't exceed the cutoff and delete self-loops
             adj = sp.csr_matrix(dist <= self.cutoff) - sp.eye(
-                n_atoms, dtype=np.bool
+                n_atoms, dtype=np.bool_
             )
             adj = adj.tocoo()
             u, v = torch.tensor(adj.row), torch.tensor(adj.col)

examples/pytorch/grace/eval.py

@@ -34,7 +34,7 @@ def repeat(n_times):
 
 
 def prob_to_one_hot(y_pred):
-    ret = np.zeros(y_pred.shape, np.bool)
+    ret = np.zeros(y_pred.shape, np.bool_)
     indices = np.argmax(y_pred, axis=1)
     for i in range(y_pred.shape[0]):
         ret[i][indices[i]] = True
@@ -61,7 +61,7 @@ def label_classification(
     Y = y.detach().cpu().numpy()
     Y = Y.reshape(-1, 1)
     onehot_encoder = OneHotEncoder(categories="auto").fit(Y)
-    Y = onehot_encoder.transform(Y).toarray().astype(np.bool)
+    Y = onehot_encoder.transform(Y).toarray().astype(np.bool_)
 
     X = normalize(X, norm="l2")
 

examples/pytorch/graphsaint/utils.py

@@ -70,14 +70,14 @@ def load_data(args, multilabel):
     DataType = namedtuple("Dataset", ["num_classes", "train_nid", "g"])
 
     adj_full = scipy.sparse.load_npz("./{}/adj_full.npz".format(prefix)).astype(
-        np.bool
+        np.bool_
     )
     g = dgl.from_scipy(adj_full)
     num_nodes = g.num_nodes()
 
     adj_train = scipy.sparse.load_npz(
         "./{}/adj_train.npz".format(prefix)
-    ).astype(np.bool)
+    ).astype(np.bool_)
     train_nid = np.array(list(set(adj_train.nonzero()[0])))
 
     role = json.load(open("./{}/role.json".format(prefix)))

examples/pytorch/model_zoo/geometric/coarsening.py

@@ -157,7 +157,7 @@ def HEM_one_level(rr, cc, vv, rid, weights):
     nnz = rr.shape[0]
     N = rr[nnz - 1] + 1
 
-    marked = np.zeros(N, np.bool)
+    marked = np.zeros(N, np.bool_)
     rowstart = np.zeros(N, np.int32)
     rowlength = np.zeros(N, np.int32)
     cluster_id = np.zeros(N, np.int32)

examples/pytorch/pinsage/data_utils.py

@@ -11,9 +11,9 @@ import tqdm
 # takes.  It essentially follows the intuition of "training on the past and predict the future".
 # One can also change the threshold to make validation and test set take larger proportions.
 def train_test_split_by_time(df, timestamp, user):
-    df["train_mask"] = np.ones((len(df),), dtype=np.bool)
-    df["val_mask"] = np.zeros((len(df),), dtype=np.bool)
-    df["test_mask"] = np.zeros((len(df),), dtype=np.bool)
+    df["train_mask"] = np.ones((len(df),), dtype=np.bool_)
+    df["val_mask"] = np.zeros((len(df),), dtype=np.bool_)
+    df["test_mask"] = np.zeros((len(df),), dtype=np.bool_)
     df = dd.from_pandas(df, npartitions=10)
 
     def train_test_split(df):