`get_vector_length` incorrectly returns for shared variable without static shape

[ricardoV94]

Description

import pytensor
import pytensor.tensor as pt
pt.get_vector_length(pytensor.shared(np.zeros(3,)))  # 3

It should raise because the variable has type.shape=(None,)

[Abhinav-Khot]

@ricardoV94 This issue arises due to this registration in tensor/sharedvar.py

@_get_vector_length.register(TensorSharedVariable)
def _get_vector_length_TensorSharedVariable(var_inst, var):
    return len(var.get_value(borrow=True))

How should I move forward? Should I remove this registration (which will cause the snipped you have provided to raise an error)? Or should it be left as is?

[ricardoV94]

Yes we should remove that but make sure that if the shape of the shared variable is static it still works. I suspect we don't need to do anything and and the dispatch on the base class handles that case

[Abhinav-Khot]

@ricardoV94 Should not be an issue, the implementation of get_vector_length first checks if the shape is static before calling the dispatches.

def get_vector_length(v: TensorLike) -> int:
    """Return the run-time length of a symbolic vector, when possible.

    Parameters
    ----------
    v
        A rank-1 `TensorType` variable.

    Raises
    ------
    TypeError
        `v` hasn't the proper type.
    ValueError
        No special case applies, the length is not known.
        In general this is not possible, but for a number of special cases
        the length can be determined at compile / graph-construction time.
        This function implements these special cases.

    """
    v = as_tensor_variable(v)
    
    if v.type.ndim != 1:
        raise TypeError(f"Argument must be a vector; got {v.type}")

    static_shape: int | None = v.type.shape[0]
    if static_shape is not None:
        return static_shape

    return _get_vector_length(getattr(v.owner, "op", v), v)

A small test for the same:

import pytensor
import numpy as np
from pytensor import tensor as pt

val = np.zeros(3, dtype='float32')
x = pytensor.shared(val, name='x', shape = val.shape, strict=True)
print(pt.get_vector_length(x)) # returns 3

pt.get_vector_length(pytensor.shared(val))# Raises ValueError : "ValueError: Length of <Vector(float32, shape=(?,))> cannot be determined"