TypeError: iteration over a 0-d array, using numpy

Question

I'm using this code, straight from the scikit-learn page. It creates a dictionary:

symbol_dict = {
    'TOT': 'Total',
    'XOM': 'Exxon',
    'CVX': 'Chevron',}
symbols, names = np.array(symbol_dict.items()).T

But I get an error:

TypeError: iteration over a 0-d array

This code is straight from the example code, so I have no idea what's going wrong.

Your code works perfectly when I try it. The problem seems to be coming from something you haven't shown. The traceback would make debugging this much easier. — user2357112, Jan 18 '14 at 00:52
Are you on Python 3? On Python 3, `items` returns a dict view, which NumPy doesn't understand. Use `np.array(list(symbol_dict.items())).T` to turn the view into a list of tuples, which NumPy will recognize properly. — user2357112, Jan 18 '14 at 01:03
user2357112, This fixed it for me too. Why not post this as the answer? — Manuel Hernandez, Jul 03 '16 at 13:47

score 2 · Answer 1 · answered Dec 15 '17 at 23:03

As user2357112 said, in Python 3 dict.items() returns a dictionary view object rather than a list of key-value pairs (the difference is explained here). Wrapping it in list() creates a list which is something NumPy can turn into an array:

np.array(list(symbol_dict.items()))

TypeError: iteration over a 0-d array, using numpy

1 Answers1