多项式与线性回归

# coding: utf-8

# In[15]:

import  numpy as  np 
from  sklearn.pipeline import Pipeline
from sklearn.linear_model import LinearRegression
from sklearn.preprocessing import PolynomialFeatures

# In[16]:

n_dots = 200
X =  np.linspace(-2*np.pi,2*np.pi,n_dots)
Y = np.sin(X) + 0.2* np.random.rand(n_dots) -0.1
X = X.reshape(-1,1)
Y = Y.reshape(-1,1)

# In[20]:

def poly_model(degree = 1):
    poly = PolynomialFeatures(degree = degree)
    lr = LinearRegression()
    pipeline = Pipeline([
        ("poly",poly),
        ("lr",lr)
    ])
    
    return pipeline

# In[21]:

from sklearn.metrics import mean_squared_error

# In[24]:

degrees = [2,3,5,10]
results = []
for i in degrees:
    model = poly_model(degree=i)
    model.fit(X,Y)
    train_score = model.score(X,Y)
    mse = mean_squared_error(Y,model.predict(X))
    results.append({"model":model,"degree":i,"score":train_score,"mse":mse})
    
    
for r in results:
    print("degree:{};train_score:{};mean_squared_score:{}".format(r["degree"],r["score"],r["mse"]))

# In[26]:

import matplotlib.pyplot as plt

# In[32]:

plt.figure(figsize=(12,6))
for i ,r in enumerate(results):
    fig = plt.subplot(2,2,i+1)
    plt.xlim(-8,8)
    plt.title("LineRegression degree = {}".format(r["degree"]))
    plt.scatter(X,Y,s = 5,c ='b',alpha=0.9)
    plt.plot(X,r["model"].predict(X),'r-')