test_v1.1

encoders/__init__.py

@@ -1,6 +0,0 @@
-from tugo.encoders.base import *
-#from tugo.encoders.alphago import *
-#from tugo.encoders.betago import *
-from tugo.encoders.oneplane import *
-from tugo.encoders.sevenplane import *
-from tugo.encoders.simple import *

encoders/alphago.py

@@ -134,8 +134,3 @@ class AlphaGoEncoder(Encoder):
 
     def shape(self):
         return self.num_planes, self.board_height, self.board_width
-
-
-def create(board_size):
-    return AlphaGoEncoder(board_size)
-

encoders/alphago_test.py

@@ -1,26 +0,0 @@
-import unittest
-
-from tugo.agent.helpers import is_point_an_eye
-from tugo.goboard_fast import Board, GameState, Move
-from tugo.gotypes import Player, Point
-from tugo.encoders.alphago import AlphaGoEncoder
-
-
-class AlphaGoEncoderTest(unittest.TestCase):
-    def test_encoder(self):
-        alphago = AlphaGoEncoder()
-
-        start = GameState.new_game(19)
-        next_state = start.apply_move(Move.play(Point(16, 16)))
-        alphago.encode(next_state)
-
-        self.assertEquals(alphago.name(), 'alphago')
-        self.assertEquals(alphago.board_height, 19)
-        self.assertEquals(alphago.board_width, 19)
-        self.assertEquals(alphago.num_planes, 49)
-        self.assertEquals(alphago.shape(), (49, 19, 19))
-
-
-
-if __name__ == '__main__':
-    unittest.main()

encoders/base.py

@@ -1,12 +1,4 @@
-# tag::importlib[]
 import importlib
-# end::importlib[]
-
-__all__ = [
-    'Encoder',
-    'get_encoder_by_name',
-]
-
 
 # tag::base_encoder[]
 class Encoder:

encoders/oneplane.py

@@ -1,9 +1,8 @@
-# tag::oneplane_imports[]
 import numpy as np
 
 from tugo.encoders.base import Encoder
 from tugo.goboard import Point
-# end::oneplane_imports[]
+
 
 
 # tag::oneplane_encoder[]

gosgf/__init__.py

@@ -1 +0,0 @@
-from .sgf import *

networks/__init__.py

@@ -1,6 +1 @@
-#from .fullyconnected import *
-#from .large import *
-#from .leaky import *
-#from .medium import *
-#from .small import *
 from .alphago import *

networks/fullyconnected.py

@@ -1,14 +0,0 @@
-from __future__ import absolute_import
-from keras.layers.core import Dense, Activation, Flatten
-
-
-def layers(input_shape):
-    return [
-        Dense(128, input_shape=input_shape),
-        Activation('relu'),
-        Dense(128, input_shape=input_shape),
-        Activation('relu'),
-        Flatten(),
-        Dense(128),
-        Activation('relu'),
-    ]

networks/large.py

@@ -1,39 +0,0 @@
-from __future__ import absolute_import
-from keras.layers.core import Dense, Activation, Flatten
-from keras.layers.convolutional import Conv2D, ZeroPadding2D
-
-
-def layers(input_shape):
-    return [
-        ZeroPadding2D((3, 3), input_shape=input_shape, data_format='channels_first'),
-        Conv2D(64, (7, 7), padding='valid', data_format='channels_first'),
-        Activation('relu'),
-
-        ZeroPadding2D((2, 2), data_format='channels_first'),
-        Conv2D(64, (5, 5), data_format='channels_first'),
-        Activation('relu'),
-
-        ZeroPadding2D((2, 2), data_format='channels_first'),
-        Conv2D(64, (5, 5), data_format='channels_first'),
-        Activation('relu'),
-
-        ZeroPadding2D((2, 2), data_format='channels_first'),
-        Conv2D(48, (5, 5), data_format='channels_first'),
-        Activation('relu'),
-
-        ZeroPadding2D((2, 2), data_format='channels_first'),
-        Conv2D(48, (5, 5), data_format='channels_first'),
-        Activation('relu'),
-
-        ZeroPadding2D((2, 2), data_format='channels_first'),
-        Conv2D(32, (5, 5), data_format='channels_first'),
-        Activation('relu'),
-
-        ZeroPadding2D((2, 2), data_format='channels_first'),
-        Conv2D(32, (5, 5), data_format='channels_first'),
-        Activation('relu'),
-
-        Flatten(),
-        Dense(1024),
-        Activation('relu'),
-    ]

networks/leaky.py

@@ -1,40 +0,0 @@
-from __future__ import absolute_import
-from keras.layers import LeakyReLU
-from keras.layers.core import Dense, Flatten
-from keras.layers.convolutional import Conv2D, ZeroPadding2D
-
-
-def layers(input_shape):
-    return [
-        ZeroPadding2D((3, 3), input_shape=input_shape, data_format='channels_first'),
-        Conv2D(64, (7, 7), padding='valid', data_format='channels_first'),
-        LeakyReLU(),
-
-        ZeroPadding2D((2, 2), data_format='channels_first'),
-        Conv2D(64, (5, 5), data_format='channels_first'),
-        LeakyReLU(),
-
-        ZeroPadding2D((2, 2), data_format='channels_first'),
-        Conv2D(64, (5, 5), data_format='channels_first'),
-        LeakyReLU(),
-
-        ZeroPadding2D((2, 2), data_format='channels_first'),
-        Conv2D(48, (5, 5), data_format='channels_first'),
-        LeakyReLU(),
-
-        ZeroPadding2D((2, 2), data_format='channels_first'),
-        Conv2D(48, (5, 5), data_format='channels_first'),
-        LeakyReLU(),
-
-        ZeroPadding2D((2, 2), data_format='channels_first'),
-        Conv2D(32, (5, 5), data_format='channels_first'),
-        LeakyReLU(),
-
-        ZeroPadding2D((2, 2), data_format='channels_first'),
-        Conv2D(32, (5, 5), data_format='channels_first'),
-        LeakyReLU(),
-
-        Flatten(),
-        Dense(1024),
-        LeakyReLU(),
-    ]

networks/medium.py

@@ -1,31 +0,0 @@
-from __future__ import absolute_import
-from keras.layers.core import Dense, Activation, Flatten
-from keras.layers.convolutional import Conv2D, ZeroPadding2D
-
-
-def layers(input_shape):
-    return [
-        ZeroPadding2D((2, 2), input_shape=input_shape, data_format='channels_first'),
-        Conv2D(64, (5, 5), padding='valid', data_format='channels_first'),
-        Activation('relu'),
-
-        ZeroPadding2D((2, 2), data_format='channels_first'),
-        Conv2D(64, (5, 5), data_format='channels_first'),
-        Activation('relu'),
-
-        ZeroPadding2D((1, 1), data_format='channels_first'),
-        Conv2D(64, (3, 3), data_format='channels_first'),
-        Activation('relu'),
-
-        ZeroPadding2D((1, 1), data_format='channels_first'),
-        Conv2D(64, (3, 3), data_format='channels_first'),
-        Activation('relu'),
-
-        ZeroPadding2D((1, 1), data_format='channels_first'),
-        Conv2D(64, (3, 3), data_format='channels_first'),
-        Activation('relu'),
-
-        Flatten(),
-        Dense(512),
-        Activation('relu'),
-    ]

networks/small.py

@@ -1,33 +0,0 @@
-from __future__ import absolute_import
-
-# tag::small_network[]
-from keras.layers.core import Dense, Activation, Flatten
-from keras.layers.convolutional import Conv2D, ZeroPadding2D
-
-
-def layers(input_shape):
-    return [
-        ZeroPadding2D(padding=3, input_shape=input_shape, data_format='channels_first'),  # <1>
-        Conv2D(48, (7, 7), data_format='channels_first'),
-        Activation('relu'),
-
-        ZeroPadding2D(padding=2, data_format='channels_first'),  # <2>
-        Conv2D(32, (5, 5), data_format='channels_first'),
-        Activation('relu'),
-
-        ZeroPadding2D(padding=2, data_format='channels_first'),
-        Conv2D(32, (5, 5), data_format='channels_first'),
-        Activation('relu'),
-
-        ZeroPadding2D(padding=2, data_format='channels_first'),
-        Conv2D(32, (5, 5), data_format='channels_first'),
-        Activation('relu'),
-
-        Flatten(),
-        Dense(512),
-        Activation('relu'),
-    ]
-
-# <1> We use zero padding layers to enlarge input images.
-# <2> By using `channels_first` we specify that the input plane dimension for our features comes first.
-# end::small_network[]