Search code examples
pythonreturn-valuefunction-declaration

Python - Function has a list as argument. How to return another list without changing the first?


I'm pretty new in Python (and programming as a whole). I'm pretty sure the answer to this is obvious, but I really don't know what to do.

def do_play(value, slot, board):
    temp=board
    (i,j) = slot
    temp[i][j] = value
    return temp

board is a list of lists. value is an integer. slot is and integer tuple.

What I am trying to do here is to

  • feed the function board
  • copy board to a new list called temp
  • insert a new value in a specific location in temp
  • return temp, leaving board unchanged

When I run this is the shell, both the the original list (board) and the new list (temp) change. = \

Any help would be appreciated.


Solution

  • temp=board does not make a new board. It makes the temp variable reference the very same list as board. So changing temp[i][j] changes board[i][j] too.

    To make a copy, use

    import copy
    temp=copy.deepcopy(board)
    

    Note that temp=board[:] makes temp refer to a new list (different than board, but the contents (that is, the lists within the list) are still the same:

    In [158]: board=[[1,2],[3,4]]    
    In [159]: temp=board[:]    
    

    Modifying temp modifies board too:

    In [161]: temp[1][0]=100    
    In [162]: temp
    Out[162]: [[1, 2], [100, 4]]    
    In [163]: board
    Out[163]: [[1, 2], [100, 4]]
    

    id shows the object's memory address. This shows temp and board are different lists:

    In [172]: id(temp)
    Out[172]: 176446508
    
    In [173]: id(board)
    Out[173]: 178068780   # The ids don't match
    

    But this shows the second list inside temp is the very same list inside board:

    In [174]: id(temp[1])
    Out[174]: 178827948
    
    In [175]: id(board[1])
    Out[175]: 178827948    # The ids are the same
    

    But if you use copy.deepcopy, then the lists within the list also get copied, which is what you need if modifying temp is to leave board unchanged:

    In [164]: import copy    
    In [165]: board=[[1,2],[3,4]]    
    In [166]: temp=copy.deepcopy(board)    
    In [167]: temp[1][0]=100    
    In [168]: temp
    Out[168]: [[1, 2], [100, 4]]    
    In [169]: board
    Out[169]: [[1, 2], [3, 4]]