import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data

mnist = input_data.read_data_sets('/tmp/data/', one_hot=True)
learning_rate = 0.01
training_epochs = 25
batch_size = 100
display_step = 1

x = tf.placeholder(tf.float32, [None, 784])
y = tf.placeholder(tf.float32, [None, 10])

W = tf.Variable(tf.zeros([784, 10]))
b = tf.Variable(tf.zeros([10]))

pred = tf.nn.softmax(tf.matmul(x, W) + b)

cost = tf.reduce_mean(-tf.reduce_sum(y * tf.log(pred), reduction_indices=1))
optimizer = tf.train.GradientDescentOptimizer(learning_rate).minimize(cost)
init = tf.global_variables_initializer()

with tf.Session() as sess:
    sess.run(init)
    for epoch in range(training_epochs):
        avg_cost = 0.
        total_batch = int(mnist.train.num_examples / batch_size)
        for i in range(total_batch):
            batch_xs, batch_ys = mnist.train.next_batch(batch_size)
            _, c = sess.run([optimizer,cost], feed_dict={x: batch_xs, y: batch_ys})
            avg_cost += c / total_batch
        if (epoch + 1) % display_step == 0:
            print('Epoch: %04d' % (epoch + 1), 'cost = {:.9f}'.format(avg_cost))
    print('Optimization Finished!')

    correct_predictioon = tf.equal(tf.argmax(pred, 1), tf.argmax(y, 1))
    accuracy = tf.reduce_mean(tf.cast(correct_predictioon, tf.float32))
    print("Accuracy:", accuracy.eval({x: mnist.test.images[:3000], y: mnist.test.labels[:3000]}))