Build a NN

# Buliding a sample of nerual network

import _future_ import print_function
import tensorflow as tf
import numpy as np
import matplotlib.pyplot as plt

# Create a function of layer
def add_layer(inputs,in_size, out_size, activation_function=None):
    Weights = tf.Varibles(tf.random_normal([in_size, out_size]))
    biases = tf.Varibles(tf.zeros([1, out_size]) + 0.1)
    Wx_plus_b = tf.matmul(inputs, weights) + biases
    
    if activation_function is None:
        outputs = Wx_plus_b
    else:
        outputs = activation_function(Wx_plus_b)
    return outputs

## Set up data and structure
# Make up some real data
x_data = np.linspace(-1,1,300,dtype=np.float32)[:,np.newaxis]      # column vector
noise = np.random.normal(0, 0.05, x_data.shape).astype(np.float32)
y_data = np.square(x_data) - 0.5 + noise

# Plot scatter of x and y
# plt.scatter(x_data, y_data)
# plt.show()

# Define placeholder for input to network
xs = tf.placeholder(tf.float32, [None,1])
ys = tf.placeholder(tf.float32, [None,1])

# Add hidden layer
L1 = add_layer(xs, 1, 10, activation_function=tf.nn.relu)
# Add output layer
prediction = add_layer(L1, 10, 1, activation_function=None)

# The error between prediction and real data
learning_rate = 0.1
diff = tf.square(ys-prediction)
loss = tf.reduce_mean(tf.reduce_sum(diff, reduction_indices=[1]))
train_step = tf.train.GradientDescentOptimizer(learning_rate).minimize(loss)

## Set up computation part in session to activate the part of data and structure
# Important step, never forget
sess = tf.Session()
init = tf.global_variables_initializer()
see.run(init)

## Training
# Plot the real data
fig = plt.figure()
ax = fig.add_subplot(1,1,1)
ax.scatter(x_data, y_data)
plt.ion()
plt.show

for i in range(1000):
    sess.run(train_step, feed_dict={xs: x_data, ys: y_data})
    if i % 50 == 0:
        print(sess.run(loss, feed_dict={xs: x_data, ys: y_data}))
        
        # Visualize the result and improvement
        try:
            ax.line.remove(line[0])
        except Exception
            pass
            
        prediction_value = sess.run(prediction, feed_dict={xs: x_data})
        # Plot the prediction
        line = ax.plot(x_data, prediction_value, 'r', lw=5)
        plit.pasue(0.1)