Construct a tensor of tf.Variable objects

Question

My cost function involves the matrix

T=[[1.0-a,b],[a,1.0-b]]

I can define

import numpy as np
import tensorflow as tf
a=0.3
b=0.4
T = tf.Variable([[1.0-a,b],[a,1.0-b]]

and this works well in the optimization, but then I am saying that I have four variables: 1-a,b,a,1-b (the gradient has four elements). On the other hand, I would like my variables to be two: a and b (the gradient has two elements).

I thought of doing something like

var = tf.Variable([a,b])
T = tf.constant([[1.0-var[0],var[1]],[var[0],1.0-var[1]]])

but this does not work, outputing the following:

TypeError: List of Tensors when single Tensor expected

So how can I construct a tensor made of tf.Variable objects?

Thank you.

What kind of cost function is this T=[[1.0-a,b],[a,1.0-b]]? In what sense is this a cost function? — nbro
– nbro, Commented Jun 7, 2018 at 15:24
I said my cost function involves T. To be specific, my cost function will depend on vector elements of r = T^N v, where v is just [1,1], and N is some large number. Actually, as you might infer, r is just the dominant eigenvector of T, when N is large enough. — Mencia
– Mencia, Commented Jun 7, 2018 at 15:27

javidcf · Accepted Answer · 2018-06-07 15:31:31Z

2

I think what you need is:

import tensorflow as tf

a = tf.Variable(0.3)
b = tf.Variable(0.4)
T = tf.convert_to_tensor([[1.0 - a, b], [a, 1.0 - b]]

answered Jun 7, 2018 at 15:31

javidcf

59.9k7 gold badges87 silver badges134 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

cs95 · Accepted Answer · 2018-06-07 15:31:09Z

1

Initialise T as a placeholder.

T = tf.placeholder(tf.float32, [2, 2])

When starting your session, compute T and pass it in through a feed_dict:

with tf.Session() as sess:
    a, b = .3, .4
    inp = np.array([[1 - a, b], [a, 1 - b]])
    sess.run(optimizer, feed_dict={T : inp})

Where optimizer is the node that minimizes your cost function.

answered Jun 7, 2018 at 15:31

cs95

406k106 gold badges744 silver badges797 bronze badges

1 Comment

Mencia Over a year ago

Thank you for your answer. @jdehesa answer was more helpful because I need to use tf.Variable since I am using pymanopt (pymanopt.github.io), see section c in Cost Functions.

Collectives™ on Stack Overflow

Construct a tensor of tf.Variable objects

2 Answers 2

Comments

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related