# 410621225葉映辰 深度學習作業
# 問題:
八位元(1或0)數列中，若"1"的出現次數為偶數輸出1,反之輸出0。
請用深度學習辨識輸入的八位元(1或0)數列輸出的結果。
# 解題思路:
## 第一部分.深度學習用的資料庫之搭建

def iqwe(x):
    cntoy = 1
    while(x):
        cntoy += 1
        x &= x-1
    #print(cntoy & 1)
    return cntoy & 1
    
zeros = np.zeros( (256, 8) )

for a_total in range(0,256):
    a = a_total
    for i in range(0, 8):
        if a >= 2**(7-i):
            a = a - 2**(7-i)
            zeros[a_total, i] = 1


one = np.zeros((256,1))
for i in range(0,256):
    one[i,0] = iqwe(i)

### 測試用題目:
用迴圈產生0到255的十進位數,並轉成二進位後。

### 測試用答案:
呼叫子程式計算每組數烈,"1”出現的奇偶數，以輸出1或0。

# 第二部分. 模型製作

class Linear:
   def __init__(self, m, n):
       self.W = np.random.randn(m, n)
       self.b = np.random.randn(1, n)
       self.dW = None
       self.db = None

   def forward(self, x):
       self.x = x
       out = np.dot(x, self.W) + self.b
       #print("self.W : ",self.W.T)
       #print("self.b : ",self.b.T)
       return out

   def backward(self, dout):
       dx = np.dot(dout, self.W.T)

       self.dW = np.dot(self.x.T, dout)
       self.db = np.sum(dout, axis=0)
       return dx


class ReLU:
   def __init__(self):
       pass

   def forward(self, x):
       self.mask = (x <= 0)
       out = x
       out[out <= 0] = 0
       return out

   def backward(self, dout):
       dx = dout
       dx[self.mask] = 0
       return dx


class Sigmoid:
   def __int__(self):
       pass

   def forward(self, x):
       out = 1 / (1 + np.exp(-x))
       self.o = out
       return out

   def backward(self, dout):
       dx = dout * self.o * (1 - self.o)
       return dx


class Loss:
   def __inint__(self):
       pass

   def forward(self, y, ybar):
       self.ybar = ybar
       return np.sum((y - ybar) ** 2)

   def backward(self, dout):
       dy = -(2 * (y - self.ybar))
       return dy


class twoLayer:
   def __init__(self, m, n, o):
       self.linear0 = Linear(m, n)
       self.relu0 = ReLU()
       self.linear = Linear(n, o)
       self.sigmoid = Sigmoid()
       self.loss = Loss()
       self.last_dW = 0
       self.last_db = 0
       self.last_dW0 = 0
       self.last_db0 = 0

   def forward(self, x):
       x = self.linear0.forward(x)
       x = self.relu0.forward(x)

       x = self.linear.forward(x)
       self.ybar = self.sigmoid.forward(x)

       return self.ybar

   def backward(self, y):
       self.L = self.loss.forward(y, self.ybar)
       g = self.loss.backward(1)##############################################################
       g = self.sigmoid.backward(g)
       g = self.linear.backward(g)
       g = self.relu0.backward(g)
       g = self.linear0.backward(g)

   def update(self, eta, alpha):
       #print("alpha : ",alpha)
       self.linear.W = self.linear.W - eta * self.linear.dW + alpha * self.last_dW
       self.linear.b = self.linear.b - eta * self.linear.db + alpha * self.last_db
       self.last_dW = eta * self.linear.dW
       self.last_db = eta * self.linear.db

       self.linear0.W = self.linear0.W - eta * self.linear0.dW + alpha * self.last_dW0
       self.linear0.b = self.linear0.b - eta * self.linear0.db + alpha * self.last_db0
       self.last_dW0 = eta * self.linear0.dW
       self.last_db0 = eta * self.linear0.db

X = zeros
y = one


#X = np.array([[0, 0], [0, 1], [1, 0], [1, 1]])
#y = np.array([[0], [1], [1], [0]])

model = twoLayer(8, 128, 1)
max_epochs, chk_epochs = 100000, 10000
last_dW, last_db = 0, 0
eta, alpha = 0.04, 0.02

for e in range(max_epochs):
   model.forward(X)
   model.backward(y)
   model.update(eta, alpha)

   if (e + 1) % chk_epochs == 0:
       print(model.ybar.T)
       print('Epoch %3d: loss=%.6f' % (e + 1, model.L))
       print("")