我无法理解为什么这里需要 lmplot.通常，您希望通过采用模型函数并将其拟合到数据来执行拟合.假设你想要一个高斯函数

I have problems understanding why a lmplot is needed here. Usually you want to perform a fit by taking a model function and fit it to the data. Assume you want a gaussian function

model = lambda x, A, x0, sigma, offset:  offset+A*np.exp(-((x-x0)/sigma)**2)

您可以使用 scipy.optimize.curve_fit 将其拟合到您的数据中:

you can fit it to your data with scipy.optimize.curve_fit:

popt, pcov = curve_fit(model, df["R"].values, 
                              df["EquilibriumValue"].values, p0=[1,0,2,0])

完整代码:

import pandas as pd
import numpy as np
from scipy.optimize import curve_fit
import matplotlib.pyplot as plt

df = ... # your dataframe

# plot data
plt.scatter(df["R"].values,df["EquilibriumValue"].values, label="data")

# Fitting
model = lambda x, A, x0, sigma, offset:  offset+A*np.exp(-((x-x0)/sigma)**2)
popt, pcov = curve_fit(model, df["R"].values, 
                              df["EquilibriumValue"].values, p0=[1,0,2,0])
#plot fit
x = np.linspace(df["R"].values.min(),df["R"].values.max(),250)
plt.plot(x,model(x,*popt), label="fit")

# Fitting
model2 = lambda x, sigma:  model(x,1,0,sigma,0)
popt2, pcov2 = curve_fit(model2, df["R"].values, 
                              df["EquilibriumValue"].values, p0=[2])
#plot fit2
x2 = np.linspace(df["R"].values.min(),df["R"].values.max(),250)
plt.plot(x2,model2(x2,*popt2), label="fit2")

plt.xlim(None,10)
plt.legend()
plt.show()