np.newaxis的用法，加入一个轴

import numpy as np
import tensorflow as tf
x = np.array([0, 1, 2])
y=np.array([0,1,2])[:,np.newaxis]
print(x)
print(x.shape)
print(y)
print(y.shape)

输出结果

例如下面程序中必须要使用np.newaxis

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

def add_layer(inputs, in_size, out_size, activation_function=None):
    Weights = tf.Variable(tf.random_normal([in_size, out_size]))
    biases = tf.Variable(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
x_data=np.array([0,1,2,3,4,5])[:,np.newaxis]#此处必须加np.newaxis由于  Wx_plus_b = tf.matmul(inputs, Weights) + biases
y_data=np.array([2,3,4,5,6,7])[:,np.newaxis]#此处必须加np.newaxis
xs=tf.placeholder(tf.float32,[None,1])
ys=tf.placeholder(tf.float32,[None,1])
l1=add_layer(xs,1,10,tf.nn.relu)
predict=add_layer(l1,10,1,None)
loss=tf.reduce_sum(tf.square(y_data-predict))
train_step= tf.train.GradientDescentOptimizer(0.001).minimize(loss)
session=tf.Session()
init = tf.global_variables_initializer()
session.run(init)
# plot the real data
fig = plt.figure()
ax = fig.add_subplot(1,1,1)
ax.scatter(x_data, y_data)
plt.ion()
plt.show()

MSE=[]
for i in range(10000):
    # training
    _,train_loss=session.run([train_step,loss], feed_dict={xs: x_data, ys: y_data})
    MSE.append(train_loss)
    if i % 100 == 0:
        print(train_loss)
        try:
            ax.lines.remove(lines[0])
        except Exception:
            pass
        prediction_value = session.run(predict, feed_dict={xs: x_data})
        lines = ax.plot(x_data, prediction_value, 'r-', lw=5)
        plt.pause(1)
plt.figure(2)
plt.plot(MSE)
plt.show()