add type check for each element rather than the first element (#112)

This PR does the following:
- improvement: dramatic reduce of the call to _is_batch_set
- bugfix: list(Batch()) fail; Batch(a=[torch.ones(3), torch.ones(3)]) fail;
- misc: add type check for each element rather than the first element; add test case; _create_value with torch.Tensor does not have np.object type;

test/base/test_batch.py

@@ -8,6 +8,9 @@ from tianshou.data import Batch, to_torch
 
 
 def test_batch():
+    assert list(Batch()) == []
+    batch = Batch(a=[torch.ones(3), torch.ones(3)])
+    assert torch.allclose(batch.a, torch.ones(2, 3))
     batch = Batch(obs=[0], np=np.zeros([3, 4]))
     assert batch.obs == batch["obs"]
     batch.obs = [1]

tianshou/data/batch.py

@@ -16,10 +16,10 @@ warnings.filterwarnings(
 
 def _is_batch_set(data: Any) -> bool:
     if isinstance(data, (list, tuple)):
-        if len(data) > 0 and isinstance(data[0], (dict, Batch)):
+        if len(data) > 0 and all(isinstance(e, (dict, Batch)) for e in data):
             return True
-    elif isinstance(data, np.ndarray):
-        if isinstance(data.item(0), (dict, Batch)):
+    elif isinstance(data, np.ndarray) and data.dtype == np.object:
+        if all(isinstance(e, (dict, Batch)) for e in data.tolist()):
             return True
     return False
 
@@ -43,7 +43,8 @@ def _valid_bounds(length: int, index: Union[
         return start_valid and stop_valid
 
 
-def _create_value(inst: Any, size: int) -> Union['Batch', np.ndarray]:
+def _create_value(inst: Any, size: int) -> Union[
+        'Batch', np.ndarray, torch.Tensor]:
     if isinstance(inst, np.ndarray):
         if issubclass(inst.dtype.type, (np.bool_, np.number)):
             target_type = inst.dtype.type
@@ -54,7 +55,7 @@ def _create_value(inst: Any, size: int) -> Union['Batch', np.ndarray]:
                        dtype=target_type)
     elif isinstance(inst, torch.Tensor):
         return torch.full((size, *inst.shape),
-                          fill_value=None if inst.dtype == np.object else 0,
+                          fill_value=0,
                           device=inst.device,
                           dtype=inst.dtype)
     elif isinstance(inst, (dict, Batch)):
@@ -263,18 +264,36 @@ class Batch:
                  **kwargs) -> None:
         if copy:
             batch_dict = deepcopy(batch_dict)
-        if _is_batch_set(batch_dict):
-            self.stack_(batch_dict)
-        elif isinstance(batch_dict, (dict, Batch)):
-            for k, v in batch_dict.items():
-                if isinstance(v, dict) or _is_batch_set(v):
-                    self.__dict__[k] = Batch(v)
-                else:
-                    if isinstance(v, list):
-                        v = np.array(v)
-                        if not issubclass(v.dtype.type, (np.bool_, np.number)):
-                            v = v.astype(np.object)
-                    self.__dict__[k] = v
+        if batch_dict is not None:
+            if isinstance(batch_dict, (dict, Batch)):
+                for k, v in batch_dict.items():
+                    if isinstance(v, (list, tuple, np.ndarray)):
+                        v_ = None
+                        if not isinstance(v, np.ndarray) and \
+                                all(isinstance(e, torch.Tensor) for e in v):
+                            v_ = torch.stack(v)
+                            self.__dict__[k] = v_
+                            continue
+                        else:
+                            v_ = np.asanyarray(v)
+                        if v_.dtype != np.object:
+                            v = v_  # normal data list, this is the main case
+                            if not issubclass(v.dtype.type,
+                                              (np.bool_, np.number)):
+                                v = v.astype(np.object)
+                        else:
+                            if _is_batch_set(v):
+                                v = Batch(v)  # list of dict / Batch
+                            else:
+                                # this is actually a data list with objects
+                                v = v_
+                        self.__dict__[k] = v
+                    elif isinstance(v, dict):
+                        self.__dict__[k] = Batch(v)
+                    else:
+                        self.__dict__[k] = v
+            elif _is_batch_set(batch_dict):
+                self.stack_(batch_dict)
         if len(kwargs) > 0:
             self.__init__(kwargs, copy=copy)
 
@@ -536,9 +555,9 @@ class Batch:
         values_shared = [
             [e[k] for e in batches] for k in keys_shared]
         for k, v in zip(keys_shared, values_shared):
-            if isinstance(v[0], (dict, Batch)):
+            if all(isinstance(e, (dict, Batch)) for e in v):
                 self.__dict__[k] = Batch.stack(v, axis)
-            elif isinstance(v[0], torch.Tensor):
+            elif all(isinstance(e, torch.Tensor) for e in v):
                 self.__dict__[k] = torch.stack(v, axis)
             else:
                 v = np.stack(v, axis)

tianshou/data/utils.py

@@ -37,7 +37,7 @@ def to_torch(x: Union[torch.Tensor, dict, Batch, np.ndarray],
     elif isinstance(x, (np.number, np.bool_, Number)):
         x = to_torch(np.asanyarray(x), dtype, device)
     elif isinstance(x, list) and len(x) > 0 and \
-            isinstance(x[0], (np.number, np.bool_, Number)):
+            all(isinstance(e, (np.number, np.bool_, Number)) for e in x):
         x = to_torch(np.asanyarray(x), dtype, device)
     elif isinstance(x, np.ndarray) and \
             isinstance(x.item(0), (np.number, np.bool_, Number)):