Improve np.array implementation

[jalalium]

• Ensure np.array can be called with one variable argument (previously this caused an infinite loop over the elements of the array).
• Simplify _print_NumpyArray to reduce duplication and fix Wrong use of transpose in np.array() in fortran #1241 and fix duplicate Copy of a fortran array uses unnecessary transpose #1257.
• Improve np.array implementation to correct the default order and add the ndmin parameter.
• Add tests for new code.
• Clean up test_arrays.py so failures reference the relevant issue, tuples use == instead of np.array_equal, and a more comprehensive check is carried out on returned arrays including checking the datatype and ordering.

[EmilyBourne]

@jalalium Thanks for your pull request. Please have a quick look at the review process in the developer docs to see how we organise PRs in pyccel and add the appropriate labels. Thanks!

[yguclu]

@jalalium Please add a unit test which reproduces issue #1241. This should fail before your changes.

[jalalium]

@jalalium Please add a unit test which reproduces issue #1241. This should fail before your changes.

While trying to add tests for both epyccel and pyccel, I noticed that their behavior are inconsistent. Epyccel performs an extra transposition of the fortran code. Basically, the output of the epyccel function and the output one would get from generating fortran code are inconsistent.
For the following function:

def f():
    import numpy as np
    a = np.array([[1,  2,  4,  8, 16],
             [1,  2,  4,  8, 16],
             [1,  2,  4,  8, 16],
             [1,  2,  4,  8, 16],
             [1,  2,  4,  8, 16]], order="F")
    b = np.array(a, order="F")
    return b

epyccel outputs the following

f2 = epyccel(f, language='fortran')
print(f2())

[[ 1  1  1  1  1]
 [ 2  2  2  2  2]
 [ 4  4  4  4  4]
 [ 8  8  8  8  8]
 [16 16 16 16 16]]

While the code generated in Fortran outputs the expected result.

[EmilyBourne]

I think there are 2 bugs here. I am opening issues

[EmilyBourne]

Unfortunately after looking into this I don't think you can add tests until #1128 and #1260 are merged. I am therefore marking this PR as blocked. Hopefully @ohachim can get back to PR #1128 soon so we can unblock these issues.

[EmilyBourne]

Blocked by #1258

[yguclu]

/bot run pr_tests

[pyccel-bot]

Good job ! Your PR is using all the code it added/changed.

[pyccel-bot]

Hey @pyccel/pyccel-dev ! @jalalium has just created this great new pull request! Check it out and let me know what you think!

[yguclu]

/bot run pr_tests

[yguclu]

/bot run pr_tests

[yguclu]

Looks good, but I have a couple of questions.

[EmilyBourne]

/bot run linux

[pyccel-bot]

Good job ! Your PR is using all the code it added/changed.

[pyccel-bot]

@yguclu, @jalalium has been working hard and thinks that they have now replied to or fixed all your comments. Could you take another look at the PR and see if you can approve now?