'''
用tensorflow實現遞迴迴圈網路（LSTM）
'''
from __future__ import print_function

import tensorflow as tf
from tensorflow.contrib import rnn

#匯入MINIST資料
from tensorflow.examples.tutorials.mnist import input_data
mnist = input_data.read_data_sets("/tmp/data/",one_hot=True)
'''
為了使用遞迴神經網路對影象進行分類，我們考慮每個影象
行作為一系列畫素。 因為MNIST的影象形狀是28 * 28px，我們會
為每個樣本處理28個步驟的28個序列。
'''
 

#訓練引數
learning_rate = 0.001
training_steps = 10000
batch_size = 128
display_step = 200

#神經網路引數
num_input = 28
timesteps = 28
num_hidden = 128
num_classes = 10

#tf圖表輸入
X = tf.placeholder("float",[None,timesteps,num_input])
Y = tf.placeholder("float",[None,num_classes])

#定義權重
weights = {
    'out':tf.Variable(tf.random_normal([num_hidden,num_classes]))
}
biases = {
    'out'
 
:tf.Variable(tf.random_normal([num_classes]))
}

def RNN(x,weights,biases):
    #準備資料形狀以匹配`rnn`功能需求
    #當前資料輸入形狀：（batch_size，timesteps，n_input）
    #所需形狀：形狀的'timesteps'張量列表（batch_size，n_input）
    #Unstack獲取形狀的“時間步長”張量列表（batch_size，n_input）
    x = tf.unstack(x,timesteps,1)

    #通過tensorflow定義一個lstm單元
 

    lstm_cell = rnn.BasicLSTMCell(num_hidden,forget_bias=1.0)

    #lstm輸出單元
    outputs,states = rnn.static_rnn(lstm_cell,x,dtype=tf.float32)

    #線性啟用，使用rnn內部迴圈的最後輸出
    return tf.matmul(outputs[-1],weights['out']) + biases['out']

logits = RNN(X,weights,biases)
prediction = tf.nn.softmax(logits)

#定義損失和優化器
loss_op = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits,labels=Y))
optimizer = tf.train.GradientDescentOptimizer(learning_rate=learning_rate)
train_op = optimizer.minimize(loss_op)

#評估模型（使用測試日誌，禁用退出）
correct_pred = tf.equal(tf.argmax(prediction,1),tf.argmax(Y,1))
accuracy = tf.reduce_mean(tf.cast(correct_pred,tf.float32))

#變數初始化
init = tf.global_variables_initializer()

#開始訓練
with tf.Session() as sess:
    sess.run(init)

    for step in range(1,training_steps+1):
        batch_x,batch_y = mnist.train.next_batch(batch_size)

        #重塑資料以獲得28個元素的28個序列
        batch_x = batch_x.reshape((batch_size,timesteps,num_input))

        #執行優化操作（backprop）
        sess.run(train_op,feed_dict={X:batch_x,Y:batch_y})
        if step % display_step == 0 or step ==1:
            #計算批次損失和準確性
            loss,acc = sess.run([loss_op,accuracy],feed_dict={X:batch_x,Y:batch_y})

            print("step" + str(step) + ",Minibatch Loss=" + "{:.4f}".format(loss) +
                  ",Training Accuracy=" + "{:.3f}".format(acc))

        print("優化完成")

        #計算128個mnist測試影象準確度
        test_len = 128
        test_data = mnist.test.images[:test_len].reshape((-1,timesteps,num_input))
        test_label = mnist.test.labels[:test_len]
        print("Testing Accuracy:",sess.run(accuracy,feed_dict={X:test_data,Y:test_label}))