详细介绍如何将线性函数的所有结果合并到一个图形中

Question

所以我一直在摆弄Plotly，似乎不知道如何将线性函数的所有结果合并到Plotly中的一个图形中。对于我的数据集，我有一个例子，其中自变量称为IV，因变量为字母。下面是一个数据集示例。

现在我已经解决了如何使用Sklearn为每个因变量拟合线性模型，但它只显示在单独的图中

#Data processing import
import numpy as np
import pandas as pd
    #import csv
#Visualisation import
import matplotlib.pyplot as plt
from sklearn.linear_model import LinearRegression # Regression linear model
import plotly.express as px
import plotly.graph_objects as go
sns.set()

#===================Main function==============================================
#------------------------------------------------------------------------------
plt.close("all")

#Reading the data-------------------------------------------------------------
filename="test_dataset.xls"

#Reading the input file-------------------------------------------------------
Data0=pd.read_excel (filename)

#Read number of data, number of column and column names-----------------------
number_of_data=Data0.shape[0]
number_of_I_columns=Data0.shape[1]-1 #The first column is for the X-axis=IV, to access, eg. Data0.iloc[0,0]
column_names=Data0.columns #to access, eg. Data0.columns[1]



x = Data0.IV.values.reshape(-1, 1)

# print(x)
#fit least square for each letter data---------------------------------------------
for i in range(number_of_I_columns): #for each letter data
    
    y=Data0[column_names[i+1]].values
    y=y.astype('float64')

    #Least square fitting-----------------------------------------------------
    model = LinearRegression(fit_intercept=True)
    model.fit(x,y)
    


    #Predict the letter with the fitted model----------------------------

    x_range = np.linspace(x.min(), x.max())
    y_range = model.predict(x_range.reshape(-1, 1))


    fig = go.Figure([
            go.Scatter(name = column_names[i+1], x=Data0['IV'], y=Data0[column_names[i+1]], mode='markers'),
            go.Scatter(name='Regression Fit', x=x_range, y=y_range, mode='lines')
    ])
    fig.show()

给我这些结果：

我将所有回归拟合组合到Matplotlib中的一个图中，在这里给出了所需的结果：

您可能会问，如果Matplotlib上已经有了它，为什么还要Plotly？好的，在Plotly中，我知道我可以解开并勾选我想在图表上显示的数据，这在比较某些字母的梯度时很有用。 希望有人能帮我一把。谢谢大家!

-编辑 所以我尝试将线性函数与下面的代码结合起来。然而，结果不是在同一原点上彼此重叠。相反，它在每个结果结束后加入，显示这个图，这不是我想要的

#Reading the data-------------------------------------------------------------
filename="test_dataset.xls"

#Reading the input file-------------------------------------------------------
Data0=pd.read_excel (filename)

#Read number of data, number of column and column names-----------------------
number_of_data=Data0.shape[0]
number_of_I_columns=Data0.shape[1]-1 #The first column is for the X-axis=IV, to access, eg. Data0.iloc[0,0]
column_names=Data0.columns #to access, eg. Data0.columns[1]



x = Data0.IV.values.reshape(-1, 1)
ys = []
# print(x)
#fit least square for each letter data---------------------------------------------
for i in range(number_of_I_columns): #for each letter data
    
    y=Data0[column_names[i+1]].values
    y=y.astype('float64')
#     ys.append(y)
    #Least square fitting-----------------------------------------------------
    model = LinearRegression(fit_intercept=True)
    model.fit(x,y)
    


    #Predict the letter with the fitted model----------------------------

    x_range = np.linspace(x.min(), x.max())
    y_range = model.predict(x_range.reshape(-1, 1))
    ys.append(y_range)


###  MY ATTEMPT OF COMBINING LINEAR FUNCTIONS
ys = np.array(ys)


colnames = list(column_names)
for i in range(ys.shape[0]):
#     print(ys[:,i])
    fig = go.Figure()
    fig.add_trace(go.Scatter(x = x[:,0], y=ys[:,i], name= colnames[i+1]))
fig.show()

Answer 1

问题是您正在用fig=go.Figure()重置图形

将这条线移到循环之外应该可以解决您的问题

fig = go.Figure()
for i in range(ys.shape[0]):    
    fig.add_trace(go.Scatter(x = x[:,0], y=ys[:,i], name= colnames[i+1]))
fig.show()