python 不调用第三方库实现多次多项式拟合，多项式拟合过零点，加权多项式拟合，拟合优度与相关系数

##拟合多次多项式不过零
def fit_polynomial(x, y, degree):
    n = len(x)

    # 创建一个系数向量，初始化为0
    coefficients = [0.0] * (degree + 1)

    # 创建矩阵 A 和向量 B，用于最小二乘法
    # A = np.zeros((degree + 1, degree + 1))
    # B = np.zeros(degree + 1)
    A=[]
    B=[]

    for i in range(degree + 1):
        row = []
        B.append(0)
        for j in range(degree + 1):
            row.append(0)
        A.append(row)


    # 填充 A 和 B
    for i in range(n):
        for j in range(degree + 1):
            for k in range(degree + 1):
                A[j][k] += x[i] ** (j + k)
            B[j] += y[i] * x[i] ** j

    # 使用高斯消元法求解线性方程组
    for i in range(degree + 1):
        for j in range(i + 1, degree + 1):
            factor = A[j][i] / A[i][i]
            for k in range(i, degree + 1):
                A[j][k] -= factor * A[i][k]
            B[j] -= factor * B[i]

    # 回代求解系数
    for i in range(degree, -1, -1):
        coefficients[i] = B[i]
        for j in range(i + 1, degree + 1):
            coefficients[i] -= A[i][j] * coefficients[j]
        coefficients[i] /= A[i][i]

    #以下是求解拟合优度和相关系数的公式，不需要可以注释掉。
    x_mean = sum(x) / n
    y_mean = sum(y) / n

    res=0
    sum_=0
    total=0
    for j in range(n):
        for i in range(len(coefficients)):
            sum_=sum_+coefficients[i]*(x[j]**i)
        res=res+(sum_-y[j])**2

        total = total + (y[j] - y_mean) ** 2
        sum_ = 0
    r_squared = 1 - res / total

    up = 0.0
    down1 = 0.0
    down2 = 0.0
    for i in range(len(x)):
        up = up + (x[i] - x_mean) * (y[i] - y_mean)
        down1 = down1 + (x[i] - x_mean) ** 2
        down2 = down2 + (y[i] - y_mean) ** 2
    r = up / ((down1 * down2) ** 0.5)

    return coefficients, r_squared, r

##拟合多次多项式过零
def fit_polynomial_through_origin(x, y, degree):
    n = len(x)

    # 创建一个系数向量，初始化为0
    coefficients = [0.0] * (degree + 1)

    # 创建一个矩阵 A 和向量 B，用于最小二乘法
    A=[]
    B=[]

    for i in range(degree + 1):
        row = []
        B.append(0)
        for j in range(degree + 1):
            row.append(0)
        A.append(row)



    # 填充 A 和 B
    for i in range(n):
        for j in range(degree + 1):
            for k in range(degree + 1):
                A[j][k] += x[i] ** (j + k)
            B[j] += y[i] * x[i] ** j


    # 修改 A 和 B 以确保零截距
    for i in range(degree + 1):
        A[0][i] = 0  # 将矩阵 A 的第一行设置为零
    B[0] = 0  # 将向量 B 的第一个元素设置为零

    # 使用高斯消元法解决线性系统
    for i in range(1, degree + 1):
        for j in range(i + 1, degree + 1):
            factor = A[j][i] / A[i][i]
            for k in range(i, degree + 1):
                A[j][k] -= factor * A[i][k]
            B[j] -= factor * B[i]

    # 回代求解系数
    for i in range(degree, 0, -1):
        coefficients[i] = B[i] / A[i][i]
    #以下是求解拟合优度和相关系数的公式，不需要可以注释掉。
    x_mean = sum(x) / n
    y_mean = sum(y) / n

    res=0
    sum_=0
    total=0
    for j in range(n):
        for i in range(len(coefficients)):
            sum_=sum_+coefficients[i]*(x[j]**i)
        res=res+(sum_-y[j])**2

        total = total + (y[j] - y_mean) ** 2
        sum_ = 0

    r_squared = 1 - res / total

    up=0.0
    down1=0.0
    down2=0.0
    for i in range(len(x)):
        up= up+(x[i]-x_mean)*(y[i]-y_mean)
        down1=down1+(x[i]-x_mean)**2
        down2=down2+(y[i]-y_mean)**2
    r=up/((down1*down2)**0.5)

    return coefficients,r_squared,r

# 定义加权最小二乘法拟合函数
def weighted_fit_polynomial(x, y, weights, degree):
    n = len(x)

    # 创建一个系数向量，初始化为0
    coefficients = [0.0] * (degree + 1)

    # 创建矩阵 A 和向量 B，用于最小二乘法
    A = []
    B = []

    for i in range(degree + 1):
        row = []
        B.append(0)
        for j in range(degree + 1):
            row.append(0)
        A.append(row)

    # 填充 A 和 B
    for i in range(n):
        for j in range(degree + 1):
            for k in range(degree + 1):
                A[j][k] += weights[i] * x[i] ** (j + k)
            B[j] += weights[i] * y[i] * x[i] ** j

    # 使用高斯消元法求解线性方程组
    for i in range(degree + 1):
        for j in range(i + 1, degree + 1):
            factor = A[j][i] / A[i][i]
            for k in range(i, degree + 1):
                A[j][k] -= factor * A[i][k]
            B[j] -= factor * B[i]

    # 回代求解系数
    for i in range(degree, -1, -1):
        coefficients[i] = B[i]
        for j in range(i + 1, degree + 1):
            coefficients[i] -= A[i][j] * coefficients[j]
        coefficients[i] /= A[i][i]

    return coefficients

##浓度反算
def concentration(coefficients,area):

    concent=0
    if len(coefficients) == 2:
    #     concent=(area-coefficients[0])/coefficients[1]
        concent = (area *coefficients[1]+ coefficients[0])
    else:
        for i in range(len(coefficients)):
            concent=concent+coefficients[i]*(area**i)

    return concent