def invert_dict(input_dict):
    '''(dict) -> dict

    Return an inverted copy of input_dict. Every key-value pair in input_dict
    will be a value-key pair in the result. Since multiple values in
    input_dict may be mapped to by the same key, the values in the returned
    dictionary will be held as a set.

    input_dict is unchanged.

    >>> my_dict = {'A':1, 'B':2, 'C':2}
    >>> result = invert_dict(my_dict)
    >>> result == {1:{'A'}, 2:{'B', 'C'}}
    True
    >>> my_dict == {'A':1, 'B':2, 'C':2}
    True

    REQ: Values in input_dict must be immutable data types
    (int, float, str, tuple, bool)

    '''
    # create a new dictionary to store the result
    result = {}
    # loop through every key/value pair in input_dictionary
    for next_key in input_dict:
        next_value = input_dict[next_key]
        # 2 cases:
        # this value is already in the result dictionary
        if(next_value in result):
            # append the key to this value's set
            result[next_value].add(next_key)
        # this value is not a key in the result dictionary already
        else:
            # add the value as a key, and a new set containing the key
            # as a value
            result[next_value] = {next_key}

    return result