如何在 Tensorflow 中进行切片分配 [英] How to do slice assignment in Tensorflow

问题描述

我发现 Tensorflow 提供了 scatter_update() 来为 0 维的张量切片赋值.例如，如果张量 T 是三维的，我可以将值 v[1, :, :] 赋值给 T[i, :, :].

I found that Tensorflow provides scatter_update() to assign values to the slice of a tensor in the 0 dimension. For example, if the tensor T is three dimensional, I can assign value v[1, :, :] to T[i, :, :].

a = tf.Variable(tf.zeros([10,36,36]))   
value = np.ones([1,36,36])   
d = tf.scatter_update(a,[0],value)

with tf.Session() as sess:
    sess.run(tf.initialize_all_variables())
    print a.eval()
    sess.run(d)
    print a.eval()

但是如何将值 v[1,1,:] 赋值给 T[i,j,:]?

But how to assign values v[1,1,:] to T[i,j,:]?

a = tf.Variable(tf.zeros([10,36,36]))   
value1 = np.random.randn(1,1,36)    
e = tf.scatter_update(a,[0],value1) #Error

with tf.Session() as sess:
    sess.run(tf.initialize_all_variables())
    print a.eval()
    sess.rum(e)
    print a.eval()

TF 是否提供了其他任何功能或一种简单的方法来执行此操作?

Is there any other function that TF provide or a simple way to do this?

推荐答案

我相信你需要的是 票 #206.不过目前还没有.

I believe what you need is the assign_slice_update discussed in ticket #206. It is not yet available, though.

更新:现已实施.见jdehesa的回答:https://stackoverflow.com/a/43139565/6531137

在 assign_slice_update(或 scatter_nd())可用之前，您可以构建所需行的块，其中包含您不想修改的值以及要更新的所需值，如下所示:

Until assign_slice_update (or scatter_nd()) is available, you could build a block of the desired row containing the values you don't want to modify along with the desired values to update, like so:

import tensorflow as tf

a = tf.Variable(tf.ones([10,36,36]))

i = 3
j = 5

# Gather values inside the a[i,...] block that are not on column j
idx_before = tf.concat(1, [tf.reshape(tf.tile(tf.Variable([i]), [j]), [-1, 1]), tf.reshape(tf.range(j), [-1, 1])])
values_before = tf.gather_nd(a, idx_before)
idx_after = tf.concat(1, [tf.reshape(tf.tile(tf.Variable([i]), [36-j-1]), [-1, 1]), tf.reshape(tf.range(j+1, 36), [-1, 1])])
values_after = tf.gather_nd(a, idx_after)

# Build a subset of tensor `a` with the values that should not be touched and the values to update
block = tf.concat(0, [values_before, 5*tf.ones([1, 36]), values_after])

d = tf.scatter_update(a, i, block)

with tf.Session() as sess:
    sess.run(tf.initialize_all_variables())
    sess.run(d)
    print(a.eval()[3,4:7,:]) # Print a subset of the tensor to verify

该示例生成一个 1 张量并执行 a[i,j,:] = 5.大多数复杂性在于获取我们不想修改的值，a[i,~j,:](否则 scatter_update() 将替换这些值).

The example generate a tensor of ones and performs a[i,j,:] = 5. Most of the complexity lies into getting the values that we don't want to modify, a[i,~j,:] (otherwise scatter_update() will replace those values).

如果你想按照你的要求执行 T[i,k,:] = a[1,1,:] ，你需要替换 5*tf.ones([1, 36]) 在前面的例子中由 tf.gather_nd(a, [[1, 1]]).

If you want to perform T[i,k,:] = a[1,1,:] as you asked, you need to replace 5*tf.ones([1, 36]) in the previous example by tf.gather_nd(a, [[1, 1]]).

另一种方法是为 tf.select() 从中创建所需元素的掩码并将其分配回变量，如下所示:

Another approach would be to create a mask to tf.select() the desired elements from it and assign it back to the variable, as such:

import tensorflow as tf

a = tf.Variable(tf.zeros([10,36,36]))

i = tf.Variable([3])
j = tf.Variable([5])

# Build a mask using indices to perform [i,j,:]
atleast_2d = lambda x: tf.reshape(x, [-1, 1])
indices = tf.concat(1, [atleast_2d(tf.tile(i, [36])), atleast_2d(tf.tile(j, [36])), atleast_2d(tf.range(36))])
mask = tf.cast(tf.sparse_to_dense(indices, [10, 36, 36], 1), tf.bool)

to_update = 5*tf.ones_like(a)
out = a.assign( tf.select(mask, to_update, a) ) 

with tf.Session() as sess:
    sess.run(tf.initialize_all_variables())
    sess.run(out)
    print(a.eval()[2:5,5,:])

它在内存方面的效率可能较低，因为它需要两倍的内存来处理 a-like to_update 变量，但您可以轻松地将最后一个示例修改为从 tf.select(...) 节点获取梯度保留操作.您可能也有兴趣查看其他 StackOverflow 问题:张量值的条件分配在 TensorFlow 中.

It is potentially less efficient in terms of memory since it requires twice the memory to handle the a-like to_update variable, but you could easily modify this last example to get a gradient-preserving operation from the tf.select(...) node. You might also be interested in looking at this other StackOverflow question: Conditional assignment of tensor values in TensorFlow.

那些不雅的扭曲应该被替换为对可用的适当 TensorFlow 函数的调用.

Those inelegant contortions should be replaced to a call to the proper TensorFlow function as it becomes available.

这篇关于如何在 Tensorflow 中进行切片分配的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！