def z_score_scale_np(x):
    return (x - x.mean(axis=0)) / x.std(axis=0)

X_scaled = z_score_scale_np(X.astype(float))

        __
        __