Никита Денисов's questions

在实现线性回归时，出现了一个问题：使用梯度下降，损失一次增加几个数量级。损失 - MSE，梯度下降 - 正常。我选择了加州住房数据集作为数据集。在写代码的时候，我依靠了文章。为了不和offset分开工作，我提前在特征矩阵的开头加了一列1。班级代码

class LinearRegression():

  w = None
  alpha = None

  def __init__(self, lr, E=20):
    self.lr = lr
    self.w = np.zeros(X.shape[1] + 1)
    self.E = E

  def loss(self, X, y):
    return np.sum((X @ self.w - y) ** 2) / X.shape[0]

  def grad(self, X, y):
    grad_basic = 2 * np.transpose(X) @ (X @ self.w - y) / X.shape[0]   
    assert grad_basic.shape == (X.shape[1],) , "Градиенты должны быть столбцом из k_features + 1 элементов"
    return grad_basic

  def sgd(self, X, y, E=20):
    self.loss_arr = [self.loss(X, y)]
    for _ in tqdm(range(E)):
      if abs(self.loss_arr[-1]) < 0.1:
        break
      self.w -= self.lr * self.grad(X, y)
      self.loss_arr.append(self.loss(X, y))

  def fit(self, X, y):
    self.sgd(X, y, self.E)

  def get_params(self):
    return self.w

  def get_loss(self):
    return self.loss_arr

  def predict(self, X):
    return X.dot(self.w)

这就是 loss 的行为方式：5.637, 288906709850, 3.250e+22...
同时，解析解被正确找到。它和完整的回归代码可以在notebook中找到。请告诉我哪里出错了

我正在按照 heroku 上部署页面上的说明进行操作。但是在进入这一行（git push heroku master）之后，一切都崩溃了。错误本身：

   [ERROR] Failed to execute goal org.apache.maven.plugins:maven-compiler-plugin:3.8.1:compile (default-compile) on project myfoodserver: Fatal error compiling: invalid target release: 14 -> [Help 1]

我知道问题出在 maven 编译器上，但具体在哪里呢？我的 pom.xml：

<?xml version="1.0" encoding="UTF-8"?>
<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 https://maven.apache.org/xsd/maven-4.0.0.xsd">
<modelVersion>4.0.0</modelVersion>
<parent>
    <groupId>org.springframework.boot</groupId>
    <artifactId>spring-boot-starter-parent</artifactId>
    <version>2.2.6.RELEASE</version>
    <relativePath/>
</parent>
<groupId>ru.denisov</groupId>
<artifactId>myfoodserver</artifactId>
<version>0.0.1-SNAPSHOT</version>
<name>myfoodserver</name>
<description>Server for MyFood app</description>

<properties>
    <java.version>14</java.version>
</properties>

<dependencies>
    <dependency>
        <groupId>org.springframework.boot</groupId>
        <artifactId>spring-boot-starter-web</artifactId>
    </dependency>

    <dependency>
        <groupId>org.springframework.boot</groupId>
        <artifactId>spring-boot-starter-data-jpa</artifactId>
    </dependency>

    <dependency>
        <groupId>org.postgresql</groupId>
        <artifactId>postgresql</artifactId>
        <version>42.2.12</version>
    </dependency>
    <dependency>
        <groupId>org.springframework.boot</groupId>
        <artifactId>spring-boot-starter-test</artifactId>
        <scope>test</scope>
        <exclusions>
            <exclusion>
                <groupId>org.junit.vintage</groupId>
                <artifactId>junit-vintage-engine</artifactId>
            </exclusion>
        </exclusions>
    </dependency>

    <dependency>
        <groupId>javax.xml.bind</groupId>
        <artifactId>jaxb-api</artifactId>
        <version>2.3.0</version>
    </dependency>
</dependencies>

<build>
    <plugins>
        <plugin>
            <groupId>com.heroku.sdk</groupId>
            <artifactId>heroku-maven-plugin</artifactId>
            <version>3.0.2</version>
        </plugin>
        <plugin>
            <groupId>org.springframework.boot</groupId>
            <artifactId>spring-boot-maven-plugin</artifactId>
        </plugin>
        <plugin>
            <groupId>org.apache.maven.plugins</groupId>
            <artifactId>maven-compiler-plugin</artifactId>
            <configuration>
                <source>13</source>
                <target>13</target>
            </configuration>
        </plugin>
    </plugins>
</build>

我在做这个号码。到目前为止，不是很好

class multifilter:

    def judge_half(pos, neg):
        return pos >= neg

    def judge_any(pos, neg):
        return pos >= 1

    def judge_all(pos, neg):
        return neg == 0

    def __init__(self, iterable, *funcs, judge=judge_any):
        self.l = iterable
        self.f = funcs
        self.judge = judge
        self.i = 0

    def __iter__(self):
        return self

    def __next__(self):

        pos, neg = 0, 0
        if self.i >= len(self.l) - 1:
            raise StopIteration
        for i in self.f:
            if i(self.l[self.i]) == True:
                pos += 1
            else:
                neg += 1
        if self.judge(pos, neg) == True:
            self.i += 1
            return next(self)
        else:
            self.i += 1
            return self.l[self.i]    

但它给出了：过滤器无法正常工作。我很可能没有正确理解迭代器的原理。这就是我的想法：当我写 时，for i in multifilter(iterable, funcs, judge_any):我写for i in iter(multifilter(iterable, *funcs, judge_any))和 在循环的每一次传递中，next(). 但是当我写作时会发生什么list(multifilter(iterable, funcs, judge_any))？告诉我一些要阅读或查看迭代器的内容。只是没有哈布尔！他们的抽象类只会让我感到困惑

PS我知道我被要求在上面实现它yeld，但是在我学习基础知识之前，我不想爬到野外