亚洲乱码中文字幕综合,中国熟女仑乱hd,亚洲精品乱拍国产一区二区三区,一本大道卡一卡二卡三乱码全集资源,又粗又黄又硬又爽的免费视频

對(duì)Tensorflow中的矩陣運(yùn)算函數(shù)詳解

 更新時(shí)間:2018年07月27日 09:41:31   作者:昆侖-鄭教主  
今天小編就為大家分享一篇對(duì)Tensorflow中的矩陣運(yùn)算函數(shù)詳解,具有很好的參考價(jià)值,希望對(duì)大家有所幫助。一起跟隨小編過(guò)來(lái)看看吧

tf.diag(diagonal,name=None) #生成對(duì)角矩陣

import tensorflowas tf;
diagonal=[1,1,1,1]
with tf.Session() as sess:
  print(sess.run(tf.diag(diagonal))) 
 #輸出的結(jié)果為[[1 0 0 0]
    [0 1 0 0]
    [0 0 1 0]
    [0 0 0 1]]

tf.diag_part(input,name=None) #功能與tf.diag函數(shù)相反,返回對(duì)角陣的對(duì)角元素

import tensorflow as tf;
diagonal =tf.constant([[1,0,0,0],[0,1,0,0],[0,0,1,0],[0,0,0,1]])
with tf.Session() as sess:
 print(sess.run(tf.diag_part(diagonal)))
#輸出結(jié)果為[1,1,1,1]

tf.trace(x,name=None) #求一個(gè)2維Tensor足跡,即為對(duì)角值diagonal之和

import tensorflow as tf;
diagonal =tf.constant([[1,0,0,3],[0,1,2,0],[0,1,1,0],[1,0,0,1]])
with tf.Session() as sess:
 print(sess.run(tf.trace(diagonal)))#輸出結(jié)果為4

tf.transpose(a,perm=None,name='transpose') #調(diào)換tensor的維度順序,按照列表perm的維度排列調(diào)換tensor的順序

import tensorflow as tf;
diagonal =tf.constant([[1,0,0,3],[0,1,2,0],[0,1,1,0],[1,0,0,1]])
with tf.Session() as sess:
 print(sess.run(tf.transpose(diagonal))) #輸出結(jié)果為[[1 0 0 1]
                             [0 1 1 0]
                             [0 2 1 0]
                             [3 0 0 1]]

tf.matmul(a,b,transpose_a=False,transpose_b=False,a_is_sparse=False,b_is_sparse=False,name=None) #矩陣相乘

transpose_a=False,transpose_b=False #運(yùn)算前是否轉(zhuǎn)置

a_is_sparse=False,b_is_sparse=False #a,b是否當(dāng)作系數(shù)矩陣進(jìn)行運(yùn)算

import tensorflow as tf;
A =tf.constant([1,0,0,3],shape=[2,2])
B =tf.constant([2,1,0,2],shape=[2,2])
with tf.Session() as sess:
 print(sess.run(tf.matmul(A,B)))
#輸出結(jié)果為[[2 1]
   [0 6]]

tf.matrix_determinant(input,name=None) #計(jì)算行列式

import tensorflow as tf;
A =tf.constant([1,0,0,3],shape=[2,2],dtype=tf.float32)
with tf.Session() as sess:
 print(sess.run(tf.matrix_determinant(A))) 
#輸出結(jié)果為3.0

tf.matrix_inverse(input,adjoint=None,name=None)

adjoint決定計(jì)算前是否進(jìn)行轉(zhuǎn)置

import tensorflow as tf;
A =tf.constant([1,0,0,2],shape=[2,2],dtype=tf.float64)
with tf.Session() as sess:
 print(sess.run(tf.matrix_inverse(A)))
#輸出結(jié)果為[[ 1. 0. ]
   [ 0. 0.5]]

tf.cholesky(input,name=None) #對(duì)輸入方陣cholesky分解,即為將一個(gè)對(duì)稱正定矩陣表示成一個(gè)下三角矩陣L和其轉(zhuǎn)置的乘積德分解

import tensorflow as tf;
A =tf.constant([1,0,0,2],shape=[2,2],dtype=tf.float64)
with tf.Session() as sess:
 print(sess.run(tf.cholesky(A)))
#輸出結(jié)果為[[ 1.   0.  ]
   [ 0.   1.41421356]]

以上這篇對(duì)Tensorflow中的矩陣運(yùn)算函數(shù)詳解就是小編分享給大家的全部?jī)?nèi)容了,希望能給大家一個(gè)參考,也希望大家多多支持腳本之家。

相關(guān)文章

最新評(píng)論