feat: Implement flexible ArrayLike input handling for arrays, lists, and tuples

[Spkap]

This PR enhances the flexibility and usability of osipi functions by implementing support for array-like inputs. Users can now pass Python lists and tuples, in addition to np.ndarrays, for parameters representing array data.

Background:

During the review of PR #63 (which focused on fixing initial type annotations using NDArray), @plaresmedima and @ltorres6 suggested improving usability by accepting more general array-like types (like lists and tuples), aligning better with standard NumPy practice. This PR fulfills that follow-up action.

How it Works:

1. For Array-Like Inputs (t, ca, R1, S, etc.):

   • Type hints changed from NDArray to numpy.typing.ArrayLike.
   • Functions now call np.asarray() on these inputs upon entry. This converts lists, tuples, or existing arrays into NumPy ndarrays.
   • A subsequent check ensures the resulting array has a floating-point dtype, converting if necessary (.astype(np.float64)).
   • All downstream calculations use the ensured ndarray.

2. For Scalar Inputs (k, TR, a, Ktrans, ve, r1, etc.):

   • These parameters retain their np.floating type hints.
   • Crucially, np.asarray() is not called on scalars. They are used directly.
   • NumPy's broadcasting handles operations between these scalars and the converted arrays.

Consistency Achieved:

• Uniform Input Handling: Provides a consistent mechanism (ArrayLike + np.asarray) for all parameters designed to accept array data.
• Clear Scalar Handling: Maintains a distinct approach for single-value parameters.

Testing:

• A new test suite (tests/test_array_like_inputs.py) has been added to verify:
  • Correct function output with np.ndarray, list, and tuple inputs.
  • Proper handling of float32 and float64 dtypes.
  • Appropriate TypeError or ValueError is raised for invalid inputs (e.g., strings, incorrect dimensions).
• Existing functionality tests continue to pass, ensuring no regressions.
• All tests were run using pytest and passed successfully before submission.

[Spkap]

Hi @ltorres6,

Following up on our discussion in PR #63, I've created a separate PR to implement ArrayLike support for more flexible array inputs (lists, tuples). I have also ran tests locally with pytest.

Could you please review it when you have a chance? I will address any feedback and make any necessary changes.

Thanks you!