how to solve Ivy framework: Frontend function returned non-frontend arrays: ivy.array(1.) error?

Question

While working on a pull request for Ivy machine learning framework I got this issue:

E AssertionError: Frontend function returned non-frontend arrays: ivy.array(1.)
E   Falsifying example: test_torch_kron(
E       frontend='torch',
E       backend_fw='numpy',
E       on_device='cpu',
E       dtype_and_x=(['float16', 'float16'],
E        [array(-1., dtype=float16), array(-1., dtype=float16)]),
E       fn_tree='ivy.functional.frontends.torch.kron',
E       test_flags=FrontendFunctionTestFlags(
E           num_positional_args=0,
E           with_out=False,
E           inplace=False,
E           as_variable=[False],
E           native_arrays=[False],
E           generate_frontend_arrays=False,
E       ),
E   )
E   
E   You can reproduce this example by temporarily adding @reproduce_failure('6.82.4', b'AXicY2AAAkYGCGBEYzMwAAAAXwAF') as a decorator on your test case

How could I resolve it?

To understand the issue, I looked for other questions in stackoverflow, google and the documentation of the framework

Answer 1

Frontend functions/methods are supposed to always return a frontend array which is why we use @to_ivy_arrays_and_back. So you are supposed to get this error if you forget the decorator.

This has been explained here: https://unify.ai/docs/ivy/overview/deep_dive/ivy_frontends.html#writing-frontend-functions