dataframe数据按行做线性拟合

转载：https://blog.csdn.net/m0_37324740/article/details/79529963

数据形式：

目的：
对每一行进行线性拟合，计算斜率和评估斜率的不确定性。

方法：调用python的sklearn包中的线性回归模型计算
关键步骤：将dataframe数据类型转换成矩阵。

一、如果只是单纯计算trend，即斜率

import os
import pandas as pd
import numpy as np
from sklearn.linear_model import LinearRegression
from pandas import DataFrame,Series
from sklearn import linear_model


df = pd.read_excel(r'./1.xlsx')

X = np.array([[2010,2011,2012,2013,2014,2015,2016,2017,2018,2019,2020]]).T
regr = linear_model.LinearRegression()
trend = []

for i in range(0,376):
    y = df.iloc[i,:]
    model = regr.fit(X,y)
    trend.append(model.coef_)
    
trend1 = []
for i in trend:
    for j in i:
        trend1.append(j)
trend2 = Series(trend1)    

wl_trend = pd.concat([df,trend2],axis=1)
# print(wl_trend)
wl_trend.to_excel('./2.xlsx', index=False)
print('done!')

参考资料：https://blog.csdn.net/m0_37324740/article/details/79529963

二、需要计算评估模型的参数，如斜率标准差、p值、t值等
需要调用import statsmodels.api as sm

trend = []
bse = []
pvalues = []
tvalues = []
 

for i in range(0,376):
    y = df.iloc[i,:]
    # model = regr.fit(X,y)
    model = sm.OLS(y, X).fit()
    
    trend.append(model.params)#coef_
    bse.append(model.bse)
    pvalues.append(model.pvalues)
    tvalues.append(model.tvalues)
    
trend1 = []
for i in trend:
    for j in i:
        trend1.append(j)
trend2 = Series(trend1)    
     
bse1 = []
for i in bse:
    for j in i:
        bse1.append(j)
bse2 = Series(bse1)  
        
pvalues1 = []
for i in pvalues:
    for j in i:
        pvalues1.append(j)
pvalues2 = Series(pvalues1)  

tvalues1 = []
for i in tvalues:
    for j in i:
        tvalues1.append(j)
tvalues2 = Series(tvalues1)  

wl_trend = pd.concat([df,trend2,bse2,pvalues2,tvalues2],axis=1)
# print(wl_trend)
wl_trend.to_excel('./3.xlsx', index=False)
print('done!')

model = sm.OLS(y, X).fit()
print(model.summary())