hardestSudoku_fixed = [[True, False, False, False, False, False, False, False, False],
[False, False, True, True, False, False, False, False, False],
[False, True, False, False, True, False, True, False, False],
[False, True, False, False, False, True, False, False, False],
[False, False, False, False, True, True, True, False, False],
[False, False, False, True, False, False, False, True, False],
[False, False, True, False, False, False, False, True, True],
[False, False, True, True, False, False, False, True, False],
[False, True, False, False, False, False, True, False, False]]
def one_hot(array):
targets = np.array(np.asarray(array).reshape(-1))
one_hot = np.eye(10)[targets]
return one_hot
reducer = SolvedSudoku(hardestSudoku, hardestSudoku_fixed)
nodes1 = 1
nodes2 = 1
x = tf.placeholder("float32", [1, 810], name = "reducedBoards")
y = tf.placeholder("float32", [1, 810], name = "solutions")
def nnmodel(data):
hl1 = {'weights' : tf.Variable(tf.random_normal([810, nodes1])),
'biases' : tf.Variable(tf.random_normal([nodes1]))}
hl2 = {'weights' : tf.Variable(tf.random_normal([nodes1, nodes2])),
'biases' : tf.Variable(tf.random_normal([nodes2]))}
output_layer = {'weights' : tf.Variable(tf.random_normal([nodes2, 810])),
# [2, 8, 7, 1, 6, 9, 5, 3, 4],
# [5, 2, 1, 9, 7, 4, 3, 6, 8],
# [4, 3, 8, 5, 2, 6, 9, 1, 7],
# [7, 9, 6, 3, 1, 8, 4, 5, 2]]
#
# hardestSudoku_fixed = [[True, False, False, False, False, False, False, False, False],
# [False, False, True, True, False, False, False, False, False],
# [False, True, False, False, True, False, True, False, False],
# [False, True, False, False, False, True, False, False, False],
# [False, False, False, False, True, True, True, False, False],
# [False, False, False, True, False, False, False, True, False],
# [False, False, True, False, False, False, False, True, True],
# [False, False, True, True, False, False, False, True, False],
# [False, True, False, False, False, False, True, False, False]]
reducer = SolvedSudoku(2000)
test_reducer = SolvedSudoku(2000)
numbers_to_predict = 10
batch_size = 128
sess = tf.InteractiveSession()
x = tf.placeholder("float32", [None, 810], name="reducedBoards")
y = tf.placeholder("float32", [None, numbers_to_predict * 10],
name="solutions")
x_board = tf.reshape(x, [-1, 9, 90, 1], name="x_board")
def one_hot(array):
targets = np.array(np.asarray(array).reshape(-1))
[2, 8, 7, 1, 6, 9, 5 ,3, 4],
[5, 2, 1, 9, 7, 4, 3, 6, 8],
[4, 3, 8, 5, 2, 6, 9, 1, 7],
[7, 9, 6, 3, 1, 8, 4, 5, 2]]
hardestSudoku_fixed = [[True, False, False, False, False, False, False, False, False],
[False, False, True, True, False, False, False, False, False],
[False, True, False, False, True, False, True, False, False],
[False, True, False, False, False, True, False, False, False],
[False, False, False, False, True, True, True, False, False],
[False, False, False, True, False, False, False, True, False],
[False, False, True, False, False, False, False, True, True],
[False, False, True, True, False, False, False, True, False],
[False, True, False, False, False, False, True, False, False]]
reducer = SolvedSudoku(hardestSudoku, hardestSudoku_fixed)
learning_rate = 0.01
training_iteration = 10000
batch_size = 1
display_step = 500
x = tf.placeholder("float", [None, 81], name = "reducedBoards")
y = tf.placeholder("float", name = "solutions")
W = tf.Variable(tf.zeros([81,81]))
b = tf.Variable(tf.zeros([81]))
with tf.name_scope("Wx_b") as scope:
model = tf.nn.softmax(tf.matmul(x, W) + b)
# [5, 2, 1, 9, 7, 4, 3, 6, 8],
# [4, 3, 8, 5, 2, 6, 9, 1, 7],
# [7, 9, 6, 3, 1, 8, 4, 5, 2]]
#
# hardestSudoku_fixed = [[True, False, False, False, False, False, False, False, False],
# [False, False, True, True, False, False, False, False, False],
# [False, True, False, False, True, False, True, False, False],
# [False, True, False, False, False, True, False, False, False],
# [False, False, False, False, True, True, True, False, False],
# [False, False, False, True, False, False, False, True, False],
# [False, False, True, False, False, False, False, True, True],
# [False, False, True, True, False, False, False, True, False],
# [False, True, False, False, False, False, True, False, False]]
num_sudokus = 2000
reducer = SolvedSudoku(num_sudokus)
test_reducer = SolvedSudoku(num_sudokus//2)
numbers_to_predict = 10
batch_size = 128
sess = tf.InteractiveSession()
x = tf.placeholder("float32", [None, 810], name="reducedBoards")
y = tf.placeholder("float32", [None, numbers_to_predict*10], name="solutions")
x_board = tf.reshape(x, [-1, 9, 90, 1], name="x_board")
def one_hot(array):
targets = np.array(np.asarray(array).reshape(-1))
one_hot = np.eye(10)[targets]
return one_hot
import tensorflow as tf
import numpy as np
from Sudoku import SolvedSudoku
reducer = SolvedSudoku(1)
test_reducer = SolvedSudoku(1)
numbers_to_predict = 10
batch_size = 128
sess = tf.InteractiveSession()
x = tf.placeholder("float32", [None, 810], name="reducedBoards")
y = tf.placeholder("float32", [None, numbers_to_predict * 10],
name="solutions")
x_board = tf.reshape(x, [-1, 9, 90, 1], name="x_board")
def one_hot(array):
targets = np.array(np.asarray(array).reshape(-1))
one_hot = np.eye(10)[targets]
return one_hot
def prepare_data(x, reshape_num):
ohx = one_hot(x)
x_reshaped = np.reshape(ohx, reshape_num)
x_final = np.asarray(x_reshaped)
return x_final