I am making a little game in pygame, for use in reinforcement learning (openAI gym) but I'm encountering an issue with lists/dictionaries. I presume this is to do with empty lists all pointing to the same memory address, but I do not understand why/how or how to avoid it in this scenario.
Ultimately I want to be able to use empty lists or dicts for this, I am trying to group a list of tuples of the form ((int1, int2), int3)) by their second element (int3).
I have tried using copy() when initializing the empty array but with no difference. I did not find/understand anything useful from tutorials or docs pages, so I would greatly appreciate a brief explanation of how to avoid this issue.
def works(t):
lst = [[1],[2],[3]]
lst[1].append([x for x in t if x[1]==1])
return lst
def does_not_work(t):
lst = 3*[list()]
lst[1].append([x for x in t if x[1]==1])
return lst
t = [(('a'), 0), (('b'), 1), (('c'), 1), (('d'),2)]
print('Expected Result:')
print(works(t))
print('Confusing Result:')
print(does_not_work(t))
Expected Result:
[[1], [2, [('b', 1), ('c', 1)]], [3]]
Confusing Result:
[[[('b', 1), ('c', 1)]], [[('b', 1), ('c', 1)]], [[('b', 1), ('c', 1)]]]
