Hello everyone,
I am going through Theano Tutorials
(http://deeplearning.net/software/theano/tutorial/gradients.html). I have a
basic doubt:
why the foloowing return an error :
In [28]: x = T.dmatrix('x')
...: y = x**2
...: gy = T.grad(y, x)
...: dy = theano.function([x], gy)
...:
...:
---------------------------------------------------------------------------
TypeError Traceback (most recent call last)
<ipython-input-28-74c00692575f> in <module>()
1 x = T.dmatrix('x')
2 y = x**2
----> 3 gy = T.grad(y, x)
4 dy = theano.function([x], gy)
5
py27_venvs/venv/local/lib/python2.7/site-packages/theano/gradient.pyc in
grad(cost, wrt, consider_constant, disconnected_inputs, add_names, known_grads,
return_disconnected, null_gradients)
485
486 if cost is not None and cost.ndim != 0:
--> 487 raise TypeError("cost must be a scalar.")
488
489 if isinstance(wrt, set):
while writing y = T.sum(x**2) gives me elementwise grad succesfully.
In [29]: x = T.dmatrix('x')
...: y = T.sum(x**2)
...: gy = T.grad(y, x)
...: dy = theano.function([x], gy)
...:
In [30]: dy([[0, 1], [-1, -2]])
Out[30]:
array([[ 0., 2.],
[-2., -4.]])
with regards,
Satti
--
---
You received this message because you are subscribed to the Google Groups
"theano-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/theano-users/056bddee-4db4-4324-9706-64e5bbbb2783%40googlegroups.com.
