basic environment

.gitignore

@@ -0,0 +1,5 @@
+__pycache__
+/.settings
+/.pydevproject
+/.project
+/.settings

Core/CMazeEnviroment.py

@@ -0,0 +1,52 @@
+from enum import Enum
+import numpy as np
+
+class MazeActions(Enum):
+  LEFT = (-1, 0)
+  RIGHT = (1, 0)
+  UP = (0, -1)
+  DOWN = (0, 1)
+
+class CMazeEnviroment:
+  def __init__(self, maze, pos, FOV):
+    self._maze = np.pad(np.array(maze), FOV, constant_values=(1,))
+    self._pos = np.array(pos) + FOV
+    self._fov = FOV
+
+    self._fog = np.zeros_like(self._maze)
+    self._updateFog()
+
+  def _updateFog(self):
+    y, x = self._pos
+    self._fog[
+      x - self._fov:x + self._fov + 1,
+      y - self._fov:y + self._fov + 1
+    ] = 1
+    return
+
+  def apply(self, action):
+    self._pos += action.value
+    self._updateFog()
+    return
+
+  def vision(self):
+    y, x = self._pos
+    return self._maze[
+      x - self._fov:x + self._fov + 1,
+      y - self._fov:y + self._fov + 1
+    ]
+
+  @property
+  def state(self):
+    return ((self.vision(), self._fog, ), self.score, self.done)
+
+  @property
+  def done(self):
+    y, x = self._pos
+    return 1 < self._maze[x, y]
+
+  @property
+  def score(self):
+    h, w = self._fog.shape
+    total = h * w
+    return np.count_nonzero(self._fog) / total

main.py

@@ -0,0 +1,8 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+
+def main():
+  pass
+
+if __name__ == '__main__':
+  main()

tests/Test_CMazeEnviroment.py

@@ -0,0 +1,87 @@
+from verify import expect
+from Core.CMazeEnviroment import CMazeEnviroment, MazeActions
+import numpy as np
+
+class Test_CMazeEnviroment:
+  def test_vision(self):
+    env = CMazeEnviroment(
+      maze=[
+        [0, 0, 1, 0],
+        [0, 0, 0, 0],
+        [0, 1, 0, 0],
+        [0, 1, 0, 1],
+        [0, 0, 2, 0],
+      ],
+      pos=(0, 0),
+      FOV=1
+    )
+
+    valid = np.array([
+        [1, 1, 1],
+        [1, 0, 0],
+        [1, 0, 0],
+    ])
+    expect(str(env.vision())).is_equal(str(valid))
+
+  def test_apply(self):
+    env = CMazeEnviroment(
+      maze=[
+        [0, 0, 1, 0],
+        [0, 0, 0, 0],
+        [0, 1, 0, 0],
+        [0, 1, 0, 1],
+        [0, 0, 2, 0],
+      ],
+      pos=(0, 0),
+      FOV=1
+    )
+    env.apply(MazeActions.RIGHT)
+    env.apply(MazeActions.DOWN)
+    env.apply(MazeActions.RIGHT)
+    env.apply(MazeActions.DOWN)
+    env.apply(MazeActions.RIGHT)
+
+    valid = np.array([
+        [0, 0, 1],
+        [0, 0, 1],
+        [0, 1, 1],
+    ])
+    expect(str(env.vision())).is_equal(str(valid))
+
+  def test_increasingScoreWhileExploring(self):
+    env = CMazeEnviroment(
+      maze=[
+        [0, 0, 1, 0],
+        [0, 0, 0, 0],
+        [0, 1, 0, 0],
+        [0, 1, 0, 1],
+        [0, 0, 2, 0],
+      ],
+      pos=(0, 0),
+      FOV=1
+    )
+
+    oldScore = env.score
+    env.apply(MazeActions.RIGHT)
+    newScore = env.score
+    expect(oldScore).is_less(newScore)
+
+  def test_scoreNotChanged(self):
+    env = CMazeEnviroment(
+      maze=[
+        [0, 0, 1, 0],
+        [0, 0, 0, 0],
+        [0, 1, 0, 0],
+        [0, 1, 0, 1],
+        [0, 0, 2, 0],
+      ],
+      pos=(0, 0),
+      FOV=1
+    )
+
+    env.apply(MazeActions.RIGHT)
+    oldScore = env.score
+    env.apply(MazeActions.LEFT)
+    newScore = env.score
+    expect(oldScore).is_equal(newScore)
+