Np.array([1,0]) in np.array([[1,1]]) == True?

Is this a bug?

In [31]: np.array([1,0]) in np.array([[1,1]])                                                                                          
Out[31]: True

or, by parts

In [33]: a = np.array([[0,0]])                                                                                                         

In [34]: b = np.array([1,0])                                                                                                           

In [35]: b in a                                                                                                                        
Out[35]: True

According to this reply the in operator is meaningful only if the first operand is scalar:

Another interesting question:


Unfortunate, but thanks :slight_smile:

Out of curiosity, what is the source data? That is, what do these return?


1 Like

None really, I was just learning the syntax, thus:

In [3]: print(np.array([1,0]))                                                                                                         
[1 0]

In [4]: print(np.array([[1,1]]))                                                                                                       
[[1 1]]

by the way, it is also slightly confusing that the output format of the array is not consistent with the input:

In [5]: a = np.array([[0, 0]])                                                                                                         

In [6]: for v in a : 
   ...:     print(v) 
[0 0]

In [7]: [0 0] in a                                                                                                                     
  File "<ipython-input-7-86c19ca29a5c>", line 1
    [0 0] in a
SyntaxError: invalid syntax

Try repr(). I am not sure how it is implemented in NumPy arrays but it is designed to get a consistent representation of objects.

for v in a:
    print(f'{v!r}')    # shorthand for repr() inside formatted string