机器学习15 线性回归（销售利润表.csv）调参实战代码 reshape（1，-1）

导入包

from sklearn import datasets
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression
import pandas as pd

数据准备

data = pd.read_csv('销售利润表.csv')  #更常用
print(data.shape)
data.head()

这里对比一下，引入dataset系统自带函数的代码

---

  #波士顿房价数据集
boston = datasets.load_boston() #系统自带数据

---

#不要第一列Unamed:0
del data['Unnamed: 0']
data.head()

训练集测试集划分

train_test_split? 查看参数，得到x_train, x_test, y_train, y_test = train_test_split( x, y, test_size=0.2, random_state=42)

x_train, x_test, y_train, y_test = train_test_split( data['销售价格'], data['利润'], test_size=0.2, random_state=42)

看看是否能建模

linear = LinearRegression()
linear.fit(x_train,y_train) #一维数据不能线性回归，提示reshape(-1, 1) 

结果报错了，报错代码是：1D表示1维，arry表示numpy中的数组

ValueError: Expected 2D array, got 1D array instead: array=[ 2.264 248.57 47.992 ... 26.352 7.36 273.666]. Reshape your data either using array.reshape(-1, 1) if your data has a single feature or array.reshape(1, -1) if it contains a single sample.

一维数组reshape（1，-1）变成2维

原因是我们这里是一维数组，线性回归必须要二维数组

直接x_train.reshape(1,-1)还是会报错，因为这里一维数组是Series类型，我们要把Series先转变成numpy

首先，在首行导入包 import numpy as np

法一：Series 转换成 array 再用reshape

方法是：np.array() 就可以把Series转换成array数组

#先用np.array()把Series转换成array，然后用reshape(1,-1)转换
x_train = np.array(x_train).reshape(1,-1)
x_test = np.array(x_test).reshape(1,-1)
y_train = np.array(y_train).reshape(1,-1)
y_test = np.array(y_test).reshape(1,-1)

目前运行就可以建模了，红色部分表示有效代码，其余部分自己打草稿用

调参

代码和boston那里完全一样

result = []
for inercept in [True,False]: #截距
    for norm in [True,False]:  #归一化
        linear = LinearRegression(fit_intercept=inercept,normalize=norm)
        linear.fit(x_train,y_train)
        
        
        d = {'fit_intercept':inercept,'normalize':norm,'score':linear.score(x_test,y_test),'coef':linear.coef_ ,
             'intercept':linear.intercept_}
        result.append(d)
    
result_df = pd.DataFrame(result)
result_df
        

但是报错了，有两种报错形式，是同一个意思：数组的行列反了，应该把reshape（-1,1）改成reshape（1，-1）


修改

这样就能运行了，结果为：

结果的准确性score全是负数，一是检查模型能不能调整优化，而是看看数据清洗转化是否出错，如果都没有那就不要用这个模型了，score要越接近1才越好

法二：DataFrame

直接把data['x'] 加一对[ ]变成data[['x']],查看type（x_train）就变成了Dataframe类型，后面直接调参就可以

x_train, x_test, y_train, y_test = train_test_split( data[['销售价格']], data[['利润']], test_size=0.2, random_state=42)

或者直接出来以后强制转换一次pd.DataFrame(x_train)也和加[ ]是一样的效果

全文代码：

from sklearn import datasets
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression
import pandas as pd 
import numpy as np

data = pd.read_csv('销售利润表.csv')
print(data.shape)
data.head()

#不要第一列Unamed:0
del data['Unnamed: 0']
data.head()

x_train, x_test, y_train, y_test = train_test_split( data['销售价格'], data['利润'], test_size=0.2, random_state=42)

series转换成DataFrame方法 二选一
#先用np.array()把Series转换成array，然后用reshape(1,-1)转换
x_train = np.array(x_train).reshape(-1,1)
x_test = np.array(x_test).reshape(-1,1)
y_train = np.array(y_train).reshape(-1,1)
y_test = np.array(y_test).reshape(-1,1)

#DataFrame
x_train, x_test, y_train, y_test = train_test_split( data[['销售价格']], data[['利润']], test_size=0.2, random_state=42)

result = []
for inercept in [True,False]: #截距
    for norm in [True,False]:  #归一化
        linear = LinearRegression(fit_intercept=inercept,normalize=norm)
        linear.fit(x_train,y_train)
        
        
        d = {'fit_intercept':inercept,'normalize':norm,'score':linear.score(x_test,y_test),'coef':linear.coef_ ,
             'intercept':linear.intercept_}
        result.append(d)
    
result_df = pd.DataFrame(result)
result_df