Trying to understand this Python code

Question

I'm new to python and I don ́t understand the last line of this python code. What does it mean?

import np as numpy
 def goat_door(prizedoors, guesses):
 #strategy: generate random answers, and
 #keep updating until they satisfy the rule
 #that they aren't a prizedoor or a guess
 result = np.random.randint(0, 3, prizedoors.size)
 while True:
 bad = (result == prizedoors) | (result == guesses)
 if not bad.any():
 return result
 result[bad] = np.random.randint(0, 3, bad.sum())

prizedoors and guesses are np.random.choice(2,number of simulations)

Result is an array and I don ́t know what result[bad] means.

Edit: I've just write import np as numpy

• Clev3r: I don't think I can answer this unless I have a look at the np object. Also, I'm not so sure this is correct python, unless I'm missing something entirely. bad = ( ) | ( ) will be assigned a boolean value. Booleans in python certainly don't have a .any() function. Is this some sort of pseudo-python interpretation?
• user849425: np appears to be a reference to numpy's random library. See: docs.scipy.org/doc/numpy/reference/generated/…
• three_pineapples: @clever np is referring to numpy, which explains the use or .any()
• user849425: @Clever Also, | is the bitwise-or operator which is different from the standard or (||) operator. It is not guaranteed to return a singular boolean value.
• Clev3r: @Mike really? When would | not return a boolean?

Answer 1

result is a numpy ndarray of length prizedoors.size, where each element is randomly drawn from [0, 3). For example:

>>> result = np.random.randint(0, 3, 5)
>>> result
array([1, 1, 2, 0, 1])

bad is a boolean array which is True wherever result == prizedoors or result == guesses. Probably prizedoors and guesses are boolean arrays too. In any case, bad will wind up looking something like

>>> bad
array([ True, True, True, False, True], dtype=bool)

bad.sum() counts the number of Trues:

>>> bad.sum()
4

result[bad] selects the elements of result where bad == True:

>>> result[bad]
array([1, 1, 2, 1])

and finally, the last line fills the bad values with new random values (not necessarily good values, only new ones):

>>> result[bad] = np.random.randint(0, 3, bad.sum())
>>> result
array([1, 1, 0, 0, 1])

• Alex T: Sorry to dig this topic up, but I was wondering if its ok to change bad.sum() to len(result[bad]) ?