def uniform(size=1, precision=30):
    count = np.prod(size)
    data = coin((precision, count))
    multipliers = np.full((precision, count), 2).cumprod(axis=0)
    return (data / multipliers).sum(axis=0).reshape(size)