implement larray_equiv and larray_nan_equiv

[gdementen]

larray_equal is very picky (axes in same order and no broadcasting)

[gdementen]

Maybe more arguments to LArray.equals? See #488

[alixdamman]

@gdementen In which context that would be useful?

[gdementen]

In most contexts. Users usually do not care that axes are ordered differently when comparing two arrays. So we might want to make this the default. Also, currently, you cannot compare an array with a scalar, or to another array with less axes:

>>> arr1 = ndtest(2)
>>> arr2 = full('a=a0,a1;b=b0..b2', arr1)
>>> arr2
a\b  b0  b1  b2
 a0   0   0   0
 a1   1   1   1
>>> arr2['a0'].equals(0)
False
>>> arr2.equals(arr1)
False
>>> arr2.equals(arr2.T)
False

but...

>>> arr2 == arr1
a\b    b0    b1    b2
 a0  True  True  True
 a1  True  True  True
>>> (arr2 == arr1).all()
True
>>> arr2 == arr2.T
a\b    b0    b1    b2
 a0  True  True  True
 a1  True  True  True