How to normalize the size of a tensorflow variable -

i'm writing tensorflow code want normalize variable vector after each update. following code works well:

 sess = tf.interactivesession()  y = tf.variable(tf.random_uniform([2], -0.5, 0.5))  init = tf.initialize_all_variables()  sess.run(init)   = [2, 3]  loss = tf.reduce_sum(tf.square(a - y)) y = y / tf.sqrt(tf.reduce_sum(tf.square(y)))  optimizer = tf.train.gradientdescentoptimizer(0.05) train = optimizer.minimize(loss)  step in range(100):   sess.run(train)   temp2= sess.run(y)   print(temp2)     

and gives desired answer [ 0.55469805 0.83205169], normalized vector in direction of [2,3]

however, if change

  y = tf.variable(tf.random_uniform([2], -0.5, 0.5)) 

to

  y = tf.variable(tf.random_uniform([2,2], -0.5, 0.5)) 

and

  y = y / tf.sqrt(tf.reduce_sum(tf.square(y))) 

to

  y[0] = y[0] / tf.sqrt(tf.reduce_sum(tf.square(y[0]))) 

then error says "'variable' object not support item assignment". changed loss function

  loss = tf.reduce_sum(tf.square(a - y[0])) 

can how can normalize vector column y[0] of variable type in tensorflow?

as y tensor object , cannot assign value tensor do. hence, should works on array of tensor, , after assign value following:

 yarray = y.eval()  = [2, 3]   loss = tf.reduce_sum(tf.square(a - y.eval()[0][:]))  yarray[0][:] = yarray[0][:] / tf.sqrt(tf.reduce_sum(tf.square(yarray[0][:]))).eval()  y.assign(yarray) 

in above, array of tensor using eval function. then, compute loss function, , yarray normalization. finally, assign value of yarray y.