TensorBoard-PROJECTOR-高维向量可视化

本文主要是介绍TensorBoard-PROJECTOR-高维向量可视化，希望对大家解决编程问题提供一定的参考价值，需要的开发者们随着小编来一起学习吧！

TensorBoard-PROJECTOR-高维向量可视化

PROJECTOR用于将高维向量进行可视化，通过PCA，T-SNE等方法将高维向量投影到三维坐标系。

具体操作和解释见代码和注释：

import tensorflow as tf
import mnist_inference
import osfrom tensorflow.contrib.tensorboard.plugins import projector
from tensorflow.examples.tutorials.mnist import input_databatch_size = 128
learning_rate_base = 0.8
learning_rate_decay = 0.99
training_steps = 10000
moving_average_decay = 0.99log_dir = 'log'
sprite_file = 'mnist_sprite.jpg'
meta_file = 'mnist_meta.tsv'
tensor_name = 'final_logits'#获取瓶颈层数据，即最后一层全连接层的输出
def train(mnist):with tf.variable_scope('input'):x = tf.placeholder(tf.float32,[None,784],name='x-input')y_ = tf.placeholder(tf.float32,[None,10],name='y-input')y = mnist_inference.build_net(x)global_step = tf.Variable(0,trainable=False)with tf.variable_scope('moving_average'):ema = tf.train.ExponentialMovingAverage(moving_average_decay,global_step)ema_op = ema.apply(tf.trainable_variables())with tf.variable_scope('loss_function'):loss = tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(logits=y,labels=tf.argmax(y_,1)))with tf.variable_scope('train_step'):learning_rate = tf.train.exponential_decay(learning_rate_base,global_step,mnist.train.num_examples/batch_size,learning_rate_decay,staircase=True)train_step = tf.train.GradientDescentOptimizer(learning_rate).minimize(loss,global_step=global_step)train_op = tf.group(train_step,ema_op)with tf.Session() as sess:sess.run(tf.global_variables_initializer())for i in range(training_steps):xs,ys = mnist.train.next_batch(batch_size)_,loss_value,step = sess.run([train_op,loss,global_step],feed_dict={x:xs,y_:ys})if step % 100 == 0 :print('step:{},loss:{}'.format(step,loss_value))final_result = sess.run(y,feed_dict={x:mnist.test.images})return final_resultdef visualisation(final_result):#定义一个新向量保存输出层向量的取值y = tf.Variable(final_result,name=tensor_name)#定义日志文件writersummary_writer = tf.summary.FileWriter(log_dir)#ProjectorConfig帮助生成日志文件config = projector.ProjectorConfig()#添加需要可视化的embeddingembedding = config.embeddings.add()#将需要可视化的变量与embedding绑定embedding.tensor_name = y.name#指定embedding每个点对应的标签信息，#这个是可选的，没有指定就没有标签信息embedding.metadata_path = meta_file#指定embedding每个点对应的图像，#这个文件也是可选的，没有指定就显示一个圆点embedding.sprite.image_path = sprite_file#指定sprite图中单张图片的大小embedding.sprite.single_image_dim.extend([28,28])#将projector的内容写入日志文件projector.visualize_embeddings(summary_writer,config)#初始化向量y，并将其保存到checkpoints文件中，以便于TensorBoard读取sess = tf.InteractiveSession()sess.run(tf.global_variables_initializer())saver = tf.train.Saver()saver.save(sess,os.path.join(log_dir,'model'),training_steps)summary_writer.close()def main(_):mnist = input_data.read_data_sets('MNIST_data',one_hot=True)final_result = train(mnist)visualisation(final_result)if __name__ == '__main__':tf.app.run()

生成sprite图和meta文件:

import matplotlib.pyplot as plt
import tensorflow as tf
import numpy as np
import os
from tensorflow.examples.tutorials.mnist import input_datalog_dir = './log'
sprite_file = 'mnist_sprite.jpg'
meta_file = 'mnist_meta.tsv'def create_sprite_image(images):if isinstance(images,list):images = np.array(images)#获取图像的高和宽img_h = images.shape[1]img_w = images.shape[2]#对图像数目开方，并向上取整，得到sprite图每边的图像数目num = int(np.ceil(np.sqrt(images.shape[0])))#初始化sprite图sprite_image = np.zeros([img_h*num,img_w*num])#为每个小图像赋值for i in range(num):for j in range(num):cur = i * num + jif cur < images.shape[0]:sprite_image[i*img_h:(i+1)*img_h,j*img_w:(j+1)*img_w] = images[cur]return sprite_imageif __name__ == '__main__':mnist = input_data.read_data_sets('MNIST_data',one_hot=False)#黑底白字变成白底黑字to_visualise = 1 - np.reshape(mnist.test.images,[-1,28,28])sprite_image = create_sprite_image(to_visualise)#存储展示图像path_mnist_sprite = os.path.join(log_dir,sprite_file)plt.imsave(path_mnist_sprite,sprite_image,cmap='gray')plt.imshow(sprite_image,cmap='gray')#存储每个下标对应的标签path_mnist_metadata = os.path.join(log_dir,meta_file)with open(path_mnist_metadata,'w') as f:f.write('Index\tLabel\n')for index,label in enumerate(mnist.test.labels):f.write('{}\t{}\n'.format(index,label))