tensor.rewriting.basic – Tensor Rewrites#
Tensor optimizations addressing the ops in basic.py.
Notes
There are two ways of broadcasting arrays: second(x, y) == alloc(y, broadcast_shapes(x.shape, y.shape))
The second can be more efficient because x doesn’t usually need to be computed when we only want its shape. It may also allow other rewrites that don’t try to modify x when it has multiple clients (for fear of duplicating computation).
However, the first one is easier to reason about. Knowing we have such a graph allows to do certain rewrites such as “sinking” broadcasting operations below Elemwise. The same rewrites with alloc would be more complicated as we would need to symbolically combine the shapes of each one.
As an example contrast rewriting the following two equivalent graphs
alloc(x, broadcast_shapes(x.shape, y.shape)) + alloc(y, broadcast_shapes(x.shape, y.shape)) -> x + y second(y, x) + second(x, y) -> x + y
Theano developers (mostly) preferred to use the first form during canonicalization and introduce the second form later,
via rewrites like local_fill_to_alloc, and using the alloc_like helper inside rewrites.
Many stabilize and stabilization rewrites refuse to be applied when a variable has multiple clients, so this is important.
- pytensor.tensor.rewriting.basic.alloc_like(value: TensorVariable, template: TensorVariable, fgraph: FunctionGraph, dtype=None) TensorVariable[source]#
Fill value to the same shape and dtype as the template via alloc.
- pytensor.tensor.rewriting.basic.broadcasted_by(x: TensorVariable, y: TensorVariable) bool[source]#
Check whether x would be broadcasted by y in an Elemwise operation
- Parameters:
x (TensorVariable) – The variable that may be broadcasted by y
y (TensorVariable) – The variable that may broadcast x
- Returns:
broadcasted_by
- Return type:
bool