Merge branch 'release/1.2.7'

.github/ISSUE_TEMPLATE/bug_report.md

@@ -24,7 +24,7 @@ If applicable, add screenshots to help explain your problem.
 
 **Desktop (please complete the following information):**
  - OS: [e.g. mac/linux/windows]
- - Version [e.g. 1.2.6]
+ - Version [e.g. 1.2.7]
 
 **Additional context**
 Add any other context about the problem here.

CMakeLists.txt

@@ -1,5 +1,5 @@
 cmake_minimum_required(VERSION 3.10.0)
-project(Griddly VERSION 1.2.6)
+project(Griddly VERSION 1.2.7)
 
 set(BINARY ${CMAKE_PROJECT_NAME})
 

bindings/python.cpp

@@ -12,7 +12,7 @@ namespace griddly {
 
 PYBIND11_MODULE(python_griddly, m) {
   m.doc() = "Griddly python bindings";
-  m.attr("version") = "1.2.6";
+  m.attr("version") = "1.2.7";
 
 #ifndef NDEBUG
   spdlog::set_level(spdlog::level::debug);
@@ -102,7 +102,6 @@ PYBIND11_MODULE(python_griddly, m) {
 
 
   py::enum_<ObserverType> observer_type(m, "ObserverType");
-  observer_type.value("NONE", ObserverType::NONE);
   observer_type.value("SPRITE_2D", ObserverType::SPRITE_2D);
   observer_type.value("BLOCK_2D", ObserverType::BLOCK_2D);
   observer_type.value("ISOMETRIC", ObserverType::ISOMETRIC);

docs/conf.py

@@ -22,7 +22,7 @@
 author = 'Chris Bamford'
 
 # The full version, including alpha/beta/rc tags
-release = '1.2.6'
+release = '1.2.7'
 
 
 # -- General configuration ---------------------------------------------------

docs/games/Partially_Observable_Zelda/index.rst

@@ -397,9 +397,15 @@ YAML
              Commands:
                - spawn: attack_fire
            Dst:
-             Object: [spider, _empty]
+             Object: spider
              Commands:
                - remove: true
+         - Src:
+             Object: avatar
+             Commands:
+               - spawn: attack_fire
+           Dst:
+             Object: _empty
    Objects:
      - Name: avatar
        Z: 3

docs/games/Zelda/index.rst

@@ -383,9 +383,16 @@ YAML
              Commands:
                - spawn: attack_fire
            Dst:
-             Object: [spider, _empty]
+             Object: spider
              Commands:
                - remove: true
+         - Src:
+             Object: avatar
+             Commands:
+               - spawn: attack_fire
+           Dst:
+             Object: _empty
+
    Objects:
      - Name: avatar
        Z: 3

docs/games/Zelda_Sequential/index.rst

@@ -383,9 +383,15 @@ YAML
              Commands:
                - spawn: attack_fire
            Dst:
-             Object: [spider, _empty]
+             Object: spider
              Commands:
                - remove: true
+         - Src:
+             Object: avatar
+             Commands:
+               - spawn: attack_fire
+           Dst:
+             Object: _empty
    Objects:
      - Name: avatar
        Z: 3

docs/reference/GDY/Actions/items/Trigger/Type/index.rst

@@ -0,0 +1,19 @@
+.. _#/properties/Actions/items/properties/Trigger/properties/Type:
+
+.. #/properties/Actions/items/properties/Trigger/properties/Type
+
+Type
+====
+
+:Description: The type of trigger to use.
+
+.. list-table::
+
+   * - **YAML Key**
+     - **Allowed Values**
+     - **Default Value**
+   * - ``Type``
+     - ``RANGE_BOX_AREA``, ``RANGE_BOX_BOUNDARY``
+     - ``RANGE_BOX_AREA``
+
+

docs/reference/GDY/Actions/items/Trigger/index.rst

@@ -21,7 +21,7 @@ Trigger
 
    * - **Property**
      - **Required**
-   * -  :ref:`TriggerType <#/properties/Actions/items/properties/Trigger/properties/TriggerType>` 
+   * -  :ref:`Type <#/properties/Actions/items/properties/Trigger/properties/Type>` 
      - 
    * -  :ref:`Range <#/properties/Actions/items/properties/Trigger/properties/Range>` 
      - 
@@ -30,5 +30,5 @@ Trigger
 .. toctree:: 
    :hidden:
 
-   TriggerType/index
+   Type/index
    Range/index

docs/tutorials/Proximity Triggers/index.rst

@@ -84,7 +84,7 @@ Additionally you can set a ``Probability`` for an action to set how likely the a
   - Name: set_spider_on_fire_close
     Probability: 1.0
     Trigger:
-      TriggerType: RANGE_BOX_BOUNDARY
+      Type: RANGE_BOX_BOUNDARY
       Range: 1
     Behaviours:
       - Src:
@@ -99,7 +99,7 @@ Additionally you can set a ``Probability`` for an action to set how likely the a
   - Name: set_spider_on_fire
     Probability: 0.1
     Trigger:
-      TriggerType: RANGE_BOX_BOUNDARY
+      Type: RANGE_BOX_BOUNDARY
       Range: 2
     Behaviours:
       - Src:
@@ -126,7 +126,7 @@ The action has a ``Probability`` of 0.1 of being executed.
    - Name: douse_spider
      Probability: 0.1
      Trigger:
-       TriggerType: RANGE_BOX_AREA
+       Type: RANGE_BOX_AREA
        Range: 2
      Behaviours:
        - Src:
@@ -180,7 +180,7 @@ There's a bit more boiler plate to fill out ... but otherwise proximity triggers
     - Name: set_spider_on_fire_close
       Probability: 1.0
       Trigger:
-        TriggerType: RANGE_BOX_BOUNDARY
+        Type: RANGE_BOX_BOUNDARY
         Range: 1
       Behaviours:
         - Src:
@@ -195,7 +195,7 @@ There's a bit more boiler plate to fill out ... but otherwise proximity triggers
     - Name: set_spider_on_fire
       Probability: 0.1
       Trigger:
-        TriggerType: RANGE_BOX_BOUNDARY
+        Type: RANGE_BOX_BOUNDARY
         Range: 2
       Behaviours:
         - Src:
@@ -209,7 +209,7 @@ There's a bit more boiler plate to fill out ... but otherwise proximity triggers
     - Name: douse_spider
       Probability: 0.1
       Trigger:
-        TriggerType: RANGE_BOX_AREA
+        Type: RANGE_BOX_AREA
         Range: 2
       Behaviours:
         - Src:

python/examples/rllib/rllib_multi_agent.py

@@ -2,17 +2,15 @@
 import sys
 
 import ray
-
-from examples.rllib.vector_agent import VectorAgent
-from griddly.util.rllib.callbacks import VideoCallbacks
 from ray import tune
 from ray.rllib.agents.impala import ImpalaTrainer
 from ray.rllib.models import ModelCatalog
 from ray.tune.registry import register_env
 
 from griddly import gd
-from griddly.util.rllib.torch.agents.conv_agent import SimpleConvAgent
+from griddly.util.rllib.callbacks import VideoCallbacks
 from griddly.util.rllib.environment.core import RLlibMultiAgentWrapper, RLlibEnv
+from griddly.util.rllib.torch.agents.conv_agent import SimpleConvAgent
 
 if __name__ == '__main__':
     sep = os.pathsep
@@ -22,14 +20,16 @@
 
     env_name = 'ray-ma-env'
 
+
     # Create the gridnet environment and wrap it in a multi-agent wrapper for self-play
     def _create_env(env_config):
         env = RLlibEnv(env_config)
         return RLlibMultiAgentWrapper(env, env_config)
 
+
     register_env(env_name, _create_env)
 
-    ModelCatalog.register_custom_model('SimpleConv', VectorAgent)
+    ModelCatalog.register_custom_model('SimpleConv', SimpleConvAgent)
 
     max_training_steps = 50000000
 
@@ -59,7 +59,7 @@ def _create_env(env_config):
             'yaml_file': 'Multi-Agent/robot_tag_12.yaml',
             'global_observer_type': gd.ObserverType.VECTOR,
             'player_observer_type': gd.ObserverType.VECTOR,
-            #'max_steps': 500,
+            # 'max_steps': 500,
         },
         'entropy_coeff_schedule': [
             [0, 0.01],

python/examples/rllib/rllib_single_agent.py

@@ -2,15 +2,15 @@
 import sys
 
 import ray
-from griddly.util.rllib.callbacks import VideoCallbacks
 from ray import tune
-from ray.rllib.agents.impala import ImpalaTrainer
+from ray.rllib.agents.ppo import PPOTrainer
 from ray.rllib.models import ModelCatalog
 from ray.tune.registry import register_env
 
 from griddly import gd
-from griddly.util.rllib.torch import GAPAgent
+from griddly.util.rllib.callbacks import VideoCallbacks
 from griddly.util.rllib.environment.core import RLlibEnv
+from griddly.util.rllib.torch import GAPAgent
 
 if __name__ == '__main__':
     sep = os.pathsep
@@ -29,6 +29,7 @@
         'framework': 'torch',
         'num_workers': 8,
         'num_envs_per_worker': 4,
+        'num_gpus': 1,
 
         'callbacks': VideoCallbacks,
 
@@ -63,4 +64,4 @@
         "timesteps_total": max_training_steps,
     }
 
-    result = tune.run(ImpalaTrainer, config=config, stop=stop)
+    result = tune.run(PPOTrainer, config=config, stop=stop)

python/setup.py

@@ -71,7 +71,7 @@ def griddly_package_data(config='Debug'):
 
 setup(
     name='griddly',
-    version="1.2.6",
+    version="1.2.7",
     author_email="[email protected]",
     description="Griddly Python Libraries",
     long_description=long_description,

src/Griddly/Core/Grid.cpp

@@ -342,7 +342,9 @@ std::unordered_map<uint32_t, int32_t> Grid::processCollisions() {
           std::shared_ptr<Action> collisionAction = std::shared_ptr<Action>(new Action(shared_from_this(), actionName, playerId, 0));
           collisionAction->init(object, collisionObject);
 
-          executeAndRecord(0, collisionAction);
+          auto rewards = executeAndRecord(0, collisionAction);
+
+          accumulateRewards(collisionRewards, rewards);
         }
       }
     }

tests/src/Griddly/Core/GridTest.cpp

@@ -844,17 +844,17 @@ TEST(GridTest, performActionTriggeredByCollision) {
   EXPECT_CALL(*mockObjectPtr3, isValidAction).Times(2).WillRepeatedly(Return(true));
 
   EXPECT_CALL(*mockObjectPtr1, onActionDst).Times(2).WillRepeatedly(Return(BehaviourResult{false, {{1, 1}}}));
-  EXPECT_CALL(*mockObjectPtr2, onActionDst).Times(2).WillRepeatedly(Return(BehaviourResult{false, {{1, 2}}}));
-  EXPECT_CALL(*mockObjectPtr3, onActionDst).Times(2).WillRepeatedly(Return(BehaviourResult{false, {{1, 3}}}));
+  EXPECT_CALL(*mockObjectPtr2, onActionDst).Times(2).WillRepeatedly(Return(BehaviourResult{false, {{2, 2}}}));
+  EXPECT_CALL(*mockObjectPtr3, onActionDst).Times(2).WillRepeatedly(Return(BehaviourResult{false, {{3, 3}}}));
 
   EXPECT_CALL(*mockObjectPtr1, onActionSrc(Eq("object_2"), _)).Times(1).WillOnce(Return(BehaviourResult{false, {{1, 1}}}));
   EXPECT_CALL(*mockObjectPtr1, onActionSrc(Eq("object_3"), _)).Times(1).WillOnce(Return(BehaviourResult{false, {{1, 1}}}));
 
-  EXPECT_CALL(*mockObjectPtr2, onActionSrc(Eq("object_1"), _)).Times(1).WillOnce(Return(BehaviourResult{false, {{1, 2}}}));
-  EXPECT_CALL(*mockObjectPtr2, onActionSrc(Eq("object_3"), _)).Times(1).WillOnce(Return(BehaviourResult{false, {{1, 2}}}));
+  EXPECT_CALL(*mockObjectPtr2, onActionSrc(Eq("object_1"), _)).Times(1).WillOnce(Return(BehaviourResult{false, {{2, 2}}}));
+  EXPECT_CALL(*mockObjectPtr2, onActionSrc(Eq("object_3"), _)).Times(1).WillOnce(Return(BehaviourResult{false, {{2, 2}}}));
 
-  EXPECT_CALL(*mockObjectPtr3, onActionSrc(Eq("object_2"), _)).Times(1).WillOnce(Return(BehaviourResult{false, {{1, 3}}}));
-  EXPECT_CALL(*mockObjectPtr3, onActionSrc(Eq("object_1"), _)).Times(1).WillOnce(Return(BehaviourResult{false, {{1, 3}}}));
+  EXPECT_CALL(*mockObjectPtr3, onActionSrc(Eq("object_2"), _)).Times(1).WillOnce(Return(BehaviourResult{false, {{3, 3}}}));
+  EXPECT_CALL(*mockObjectPtr3, onActionSrc(Eq("object_1"), _)).Times(1).WillOnce(Return(BehaviourResult{false, {{3, 3}}}));
 
   grid->initObject("object_1", {});
   grid->initObject("object_2", {});
@@ -874,6 +874,13 @@ TEST(GridTest, performActionTriggeredByCollision) {
       .WillOnce(Return(std::unordered_set<std::shared_ptr<Object>>{mockObjectPtr1, mockObjectPtr2, mockObjectPtr3}));
 
   auto rewards = grid->update();
+
+  ASSERT_EQ(rewards[0], 0);
+  ASSERT_EQ(rewards[1], 4);
+  ASSERT_EQ(rewards[2], 8);
+  ASSERT_EQ(rewards[3], 12);
+
+
 }
 
 }  // namespace griddly