incorrect error reported for numpy

[lhfriedman]

Steps to reproduce

1. Use this test code in file flat_iter.py

    import numpy as np
    a_0 = np.array([[1, 2, 3]])
    print(a_0)
    a_0.flat[1] = 2**14
    print(a_0)

[[1 2 3]]
[[ 1 16384 3]]

pylint flat_iter.py
************* Module flat_iter
flat_iter.py:14:0: E1137: 'a_0.flat' does not support item assignment (unsupported-assignment-operation)

---

Your code has been rated at 0.00/10

Current behavior

reports E1137

Expected behavior

ignores correct python code

pylint --version output

pylint 2.5.0
astroid 2.4.0
Python 3.7.7 (default, May 6 2020, 11:45:54) [MSC v.1916 64 bit (AMD64)]

[brycepg]

For some reason astroid is inferring the array as an ndarray. Since ndarray.flat is an iterable the above error code was produced.

import astroid

a_0_astroid = astroid.extract_node(
"""
import numpy as np
a_0 = np.array([[1, 2, 3]])
print(a_0)
a_0.flat[1] = 2**14
a_0 #@
""")
[nd_astroid] = a_0_astroid.inferred()
print(nd_astroid)

<Instance of .ndarray at 0x2253675315944>

If someone wanted to check how numpy.array is assigned in the numpy source code, that may help us understand how astroid is messing up.

[hippo91]

@brycepg ,

the problem is due to the fact that the flat attribute of numpy.ndarray is said to be None in the corresponding brain. (brain_numpy_ndarray.py).
I'll try to fix this ASAP.