对抗神经网络（python-一维数据）

 

##按照下面的代码生成的数据一直是nan，如果有大佬知道为什么，求解答

参考：https://github.com/starhou/One-dimensional-GAN/commit/797e15044e8359626ee6b6667227ff8c5b394bf8（github上一个生成一维数据的代码）

           https://www.bilibili.com/video/BV1f7411E7wU?p=4（下面的代码就是参考这个）

           https://tensorflow.google.cn/tutorials/generative/dcgan（tensorflow官网）

先保存一下，########################

from __future__ import absolute_import, division, print_function, unicode_literals
import tensorflow as tf
import glob
import imageio
import matplotlib.pyplot as plt
import numpy as np
import os
import PIL
from tensorflow.keras import layers
import time
from IPython import display
import pandas as pd
###导入我的数据
data = pd.read_excel('D://complete.xlsx') 
train_data=data.iloc[:,0:117]
train_labels=data.iloc[:,117]

train_data=train_data.astype('float32')
###归一化数据
Zmax,Zmin=train_data.max(axis=0),train_data.min(axis=0)
Z=(train_data-Zmin)/(Zmax-Zmin)
train_images = train_data.values.reshape(252,13,9,1).astype('float32')
BUFFER_SIZE = 252
BATCH_SIZE = 63
print(train_images.shape)
# 批量化和打乱数据
train_dataset = tf.data.Dataset.from_tensor_slices(train_images).shuffle(BUFFER_SIZE).batch(BATCH_SIZE)
print(train_dataset)
###生成器
def make_generator_model():
    model = tf.keras.Sequential()
    model.add(layers.Dense(128, input_shape=(100,),use_bias=False))#首次输入100个数据
    model.add(layers.BatchNormalization())
    model.add(layers.LeakyReLU())
    
    model.add(layers.Dense(160, use_bias=False))
    model.add(layers.BatchNormalization())
    model.add(layers.LeakyReLU())
    
    model.add(layers.Dense(13*9*1,use_bias=False,activation='tanh'))
    model.add(layers.BatchNormalization())
    #REshape需要以元组的形式存在
    model.add(layers.Reshape((13,9,1)))
    
    return model

###判别器
generator=make_generator_model()
def make_discriminator_model():
    model = tf.keras.Sequential()
    model.add(layers.Flatten())
    
    model.add(layers.Dense(160, use_bias=False))
    model.add(layers.BatchNormalization())
    model.add(layers.LeakyReLU())
    
    model.add(layers.Dense(128, use_bias=False))
    model.add(layers.BatchNormalization())
    model.add(layers.LeakyReLU())
    
    model.add(layers.Dense(1))

    return model
discriminator=make_discriminator_model()

cross_entropy=tf.keras.losses.BinaryCrossentropy(from_logits=True)

###判别器损失函数
def discriminator_loss(real_out,fake_out):
    real_loss=cross_entropy(tf.ones_like(real_out),real_out)
    fake_loss=cross_entropy(tf.zeros_like(fake_out),fake_out)
    return real_loss+fake_loss

###生成器损失函数
def generator_loss(fake_out):
    return cross_entropy(tf.ones_like(fake_out),fake_out)

generator_opt=tf.keras.optimizers.Adam(0.001)
discriminator_opt=tf.keras.optimizers.Adam(0.001)

EPOCHS=100
noise_dim=100

number_exp_to_generate=16
seed=tf.random.normal([number_exp_to_generate,noise_dim])

###优化生成器判别器
@tf.function
def train_step(images):
    noise = tf.random.normal([BATCH_SIZE, noise_dim])

    with tf.GradientTape() as gen_tape, tf.GradientTape() as disc_tape:
      

      real_out = discriminator(images, training=True)
      gen_image=generator(noise,training=True)
      fake_out = discriminator(gen_image, training=True)

      gen_loss = generator_loss(fake_out)
      disc_loss = discriminator_loss(real_out, fake_out)

    gradients_gen = gen_tape.gradient(gen_loss, generator.trainable_variables)
    gradients_disc = disc_tape.gradient(disc_loss, discriminator.trainable_variables)

    generator_opt.apply_gradients(zip(gradients_gen, generator.trainable_variables))
    discriminator_opt.apply_gradients(zip(gradients_disc, discriminator.trainable_variables))

###绘图
def generator_plot_images(gen_model,test_noise):
    pre_images=gen_model(test_noise,training=False)
    fig=plt.figure(figsize=(4,4))
    for i in range(pre_images.shape[0]):
        plt.subplot(4,4,i+1)
        plt.imshow((pre_images[i,:,:,0]+1)/2,cmap='gray')
    plt.show()

###按批次训练
def train(dataset,epochs):
    for epoch in range (epochs):
        for image_batch in dataset:
            train_step(image_batch)
        generator_plot_images(generator,seed)

train(train_dataset,EPOCHS)