Question

如何将类中函数返回的变量传递给该类中的另一个函数？
例如（实现线性回归）：

class ForwardPropagation(object):
    def __init__(self, xs, realoutput): 
        self.xs = np.asarray(xs)
        self.realoutput = np.asarray(realoutput)
        self.guess_slope = random.randint(1,10)
        self.guess_b = random.randint(1,10)
        self.guess_output = (self.guess_slope * self.xs) + self.guess_b

    def distance(self):#returns np.array with distance results
        abs_D = self.realoutput - self.guess_output
        D_sqr = abs_D ** 2
        return abs_D, D_sqr

    def error(self):
        sigma = np.sum(D_sqr)
        mean_error = sigma / len(D_sqr)
        return mean_error

如何将变量D_sqr（我的距离函数的平方）传递给误差函数？

Answer 1

当你可以调用其他函数时，你不需要传递任何东西

def distance(self):#returns np.array with distance results
    abs_D = self.realoutput - self.guess_output
    D_sqr = abs_D ** 2
    return abs_D, D_sqr

def error(self):
    abs_D, D_sqr = self.distance()
    sigma = np.sum(D_sqr)
    mean_error = sigma / len(D_sqr)
    return mean_error

或者你可以重新计算它

def error(self):
    abs_D = self.realoutput - self.guess_output
    D_sqr = abs_D ** 2
    sigma = np.sum(D_sqr)
    mean_error = sigma / len(D_sqr)
    return mean_error

Answer 2

使用类变量

 self.D_sqr = D_sqr

错误地说：

sigma = np.sum(self.D_sqr)

在类中的函数之间传递变量

2 个答案: