多元散射校正MSC代码

MSC

多元散射在光谱数据中产生的原因是光线在样本中传播时，与多个其他组分或颗粒发生散射。这种散射会导致光线的路径变长，使得光在样品中的传播过程中发生偏离和扩散，从而影响到光谱信号的强度和形状

输入的文件实例列是样本 行是波段

#!/usr/bin/env python3.7
# encoding: utf-8
"""
  @author: ISR
  @contact: 84692429@qq.com
  @file: 
  @time: 
  @desc:
  对采集的光谱数据进行多元散射校正（MSC）
  输入格式为csv,数据格式首行重复测量名称，首列为波段名称
"""
import numpy as np
from pandas.core.frame import DataFrame
from sklearn.linear_model import LinearRegression
import pandas as pd

def msc(data):
    """"
    data 横轴方向表示重复测量的数据，纵轴方向表示波段数量
    """
    # 计算平均光谱，实际就是x值
    s_mean = np.mean(data, axis=1)

    # 行列数
    r, c = data.shape

    # 创建一个单位矩阵
    msc_x = np.ones((r, c))

    # 遍历各列，实际是各重复测量
    for i in range(c):

        # y值
        y = data[:, i]

        # 计算光谱回归系数Ki,Bi
        lin = LinearRegression()
        lin.fit(s_mean.reshape(-1, 1), y.reshape(-1, 1))
        k = lin.coef_
        b = lin.intercept_

        msc_x[:, i] = (y - b) / k

    return msc_x

if __name__ == '__main__':

    txt_path = r"C:\Users\79420\Desktop\test\30比1\30比1.csv"
    df = pd.read_csv(txt_path, header=0, index_col=0, engine="python", encoding="utf-8")

    # 行列名称
    columns_n = df.columns.values.tolist()
    row_n = df._stat_axis.values.tolist()
    print("{0}行，{1}列".format(len(row_n),len(columns_n)))

    df_arr = df.values

    msc_arr = msc(df_arr)

    txt_path2 = r"C:\Users\79420\Desktop\test\30比1\30比1—b.csv"
    dabai1_arr_data_you = DataFrame(msc_arr, index=row_n, columns=columns_n)
    dabai1_arr_data_you.to_csv(txt_path2)

如果行是样本 列是波段 

def MSC(data):
    """
    :param data: raw spectrum data, shape (n_samples, n_features)
    :return: data after MSC: shape (n_samples, n_features)
    """
    n, p = data.shape
    msc = np.ones((n, p))

    mean = np.mean(data, axis=0) # Calculate the mean spectrum

    # Linear fitting
    for i in range(n):
        y = data[i, :]
        l = LinearRegression()
        l.fit(mean.reshape(-1, 1), y.reshape(-1, 1))
        k = l.coef_
        b = l.intercept_
        msc[i, :] = (y - b) / k

    return msc