Kernel accepts np.array as an arg only if preceded by another arg

[poojarao8]

Required prerequisites

• [x] Consult the security policy. If reporting a security vulnerability, do not report the bug using this form. Use the process described in the policy to report the issue.
• [x] Make sure you've read the documentation. Your issue may be addressed there.
• [x] Search the issue tracker to verify that this hasn't already been reported. +1 or comment there if it has.
• [ ] If possible, make a PR with a failing test to give us a starting point to work on!

Describe the bug

When I pass a numpy array as an arg to a kernel, I get the following error.

RuntimeError: error: Invalid runtime argument type. Argument of type <class 'float'> was provided, but list[float] was expected.

However, this error disappears when there is another arg preceding the np.array.

Steps to reproduce the bug

The following code gives an error:

import cudaq
import numpy as np

@cudaq.kernel
def kernel( b: np.array):
    q = cudaq.qvector(2)
    x(q[0])

bvec = np.zeros(8)
result = cudaq.sample(kernel, bvec).expectation()

Error:

RuntimeError: error: Invalid runtime argument type. Argument of type <class 'float'> was provided, but list[float] was expected.

However, the following code does NOT produce an error.

import cudaq
import numpy as np

@cudaq.kernel
def kernel( np: int, b: np.array):
    q = cudaq.qvector(2)
    x(q[0])

bvec = np.zeros(8)
result = cudaq.sample(kernel, 2, bvec).expectation()

Expected behavior

Should be able to pass a np array as an arg to a kernel.

Is this a regression? If it is, put the last known working version (or commit) here.

Not a regression

Environment

• CUDA-Q version:
• Python version:
• C++ compiler:
• Operating system:

Suggestions

No response

[sacpis]

Thanks Pooja for reporting this. Shouldn't it be np.ndarray? It works with that. np.array is a function that serves as a constructor for creating np.ndarray objects.

[poojarao8]

Good catch, it does work with np.ndarray indeed! Thank you.

[poojarao8]

Still perplexed as to why it won't raise an error with np.array when there is an arg preceding it.

[sacpis]

Sure, we will look into the second case. Unrelated to this issue, just one small thing in the example .expectation() is called on observe or observe_async.

[poojarao8]

I was just testing. It's probably not recommended, but doesn't give exp wrt Z?

[sacpis]

I was just testing. It's probably not recommended, but doesn't give exp wrt Z?

Would you please share an example?

[poojarao8]

Hey @sacpis what kind of example would you need?

[sacpis]

Hey @sacpis what kind of example would you need?

No worries.