What is the error you have?

You define x as a scalar, but you pass a matrix. This ain't work. But maybe
this isn't your original error.

Le 10 déc. 2016 21:20, "Junbin Gao" <[email protected]> a écrit :

> The following code produces index int64 error. Not sure what happened.
> Thanks heaps for your help.
>
> tl = theano.typed_list.TypedListType(theano.tensor.fmatrix)()
> l = theano.typed_list.length(tl)
> x = T.TensorType(dtype='float32', broadcastable=())('myvar')
> results, updates = theano.scan(lambda prior_result, i, b:
> T.tensordot(prior_result, b[i], [[i],[1]]),
>                                sequences=[theano.tensor.arange(2,
> dtype='int64')],
>                                non_sequences=[tl],
>                                outputs_info=[x])
> final_result = results[-1]
> tucker = theano.function(inputs=[x, b], outputs=final_result,
>                       updates=updates)
>
> X = np.array([[1, 2],[3, 4]])
> B =[np.array([[1, 2],[3, 4], [1, 3]]), np.array([[2, 1],[3, 1], [2, 3]])]
> tucker(X,B)
>
>
> --
>
> ---
> 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].
> For more options, visit https://groups.google.com/d/optout.
>

-- 

--- 
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].
For more options, visit https://groups.google.com/d/optout.

Reply via email to