Reputation: 651
In sympy, I would something like:
In [6]: import sympy as sp
In [7]: sp.var('x, y')
Out[7]: (x, y)
In [8]: X = x + y
In [9]: X.free_symbols
Out[9]: {y, x}
to get the variables X depends on. This is super convenient, because if we want to do a lambdify afterwards:
f = sp.lambdify(tuple(X.free_symbols), X)
I would like to do something similar with theano:
import theano
import theano.tensor as T
x, y = T.dvectors('x', 'y')
X = x + y
f = theano.function([x, y ], X)
But, instead of providing [x,y]
I would like to access directly the list of variables needed to create the theano.function
Is it possible ? If so, I did not find it in the theano doc, so any help or a link would be appreciated :)
Upvotes: 1
Views: 120
Reputation: 2336
Some theano functions isn't well doc-ed since mostly for internal use.
import theano
import theano.tensor as T
x, y = T.vectors('xy')
z = x+y
theano.gof.graph.inputs([z])
Output:
[x, y]
Upvotes: 3