bug fix for stacked_bar plot

[Advueu963]

TLDR: The fix comprises updating the get_n_order_values in Interactionvalues as it necessary for the stacked bar plot to work properly.

Currently the stacked_bar plot inflates the bars incorrectly and thus makes the interpretation difficult.
If given the following simple game setup

import matplotlib.pyplot as plt
import numpy as np

from shapiq import ExactComputer, powerset, Game
from shapiq.games.benchmark import SOUM

class SimpleGame(Game):
    def __init__(self) -> None:
        super().__init__(n_players=3, normalize=True, normalization_value=0)

    def value_function(self, coalitions: np.ndarray) -> np.ndarray:
        coalition_values = {
            (): 0,
            (0,): 0,
            (1,): 0,
            (2,): 0,
            (0, 1): 100,
            (0, 2): 100,
            (1, 2): 100,
            (0, 1, 2): 300,
        }

        values = np.array([coalition_values[tuple(np.where(x)[0])] for x in coalitions])

        return values

game = SimpleGame()


exact_computer = ExactComputer(n_players=game.n_players, game=game)

moebius = exact_computer("Moebius")
moebius.plot_stacked_bar(show=True)

For this game the interaction values of (0,1), (0,2) and (1,2) equal 100 and the rest is zero.
While this is correctly computed the problem lies in the visualisation:

With the fix it correctly becomes:

[mmschlk]

Ah, I see. This makes sense!

[mmschlk]

Could you document this in the CHANGELOG.md as well?

[mmschlk]

<3