from numpy.typing import ArrayLike
from spatz.metrics import Metric


class MSEMetric(Metric):
    def __init__(self) -> None:
        """Mean squared error
        """
        super().__init__()

    def _update(self, x: ArrayLike, y: ArrayLike):
        if self._score is None:
            self._score = 0.5 * (x - y)**2
        else:
            self._score += 0.5 * (x - y)**2