商业分析python实战（一）：企业所得税预测

import pandas as pd
import matplotlib
import matplotlib.pyplot as plt

#step1:获取数据
datafile = pd.read_csv('income_tax.csv',index_col='year')
print(datafile.head())

#step2:相关性分析
print(datafile.corr())

#step3:Lasso特征选择
from sklearn.linear_model import LassoCV
X = datafile.drop("y",1)   
y = datafile["y"] 
lasso = LassoCV()
lasso.fit(X, y)
coef = pd.Series(lasso.coef_, index = X.columns)
print("Lasso算法挑选了 " + str(sum(coef != 0)) + " 个变量，然后去除掉了" +  str(sum(coef == 0)) + "个变量")
imp_coef = coef.sort_values()
matplotlib.rcParams['figure.figsize'] = (8, 6)
imp_coef.plot(kind = "barh")
plt.title("Lasso Model Feature Importance")
plt.show()

#step4:GM算法进行变量灰色预测
def GM11(x0,i): 
    import numpy as np
    x1 = x0.cumsum() 
    x1 =pd.DataFrame(x1)
    z1 = (x1 + x1.shift())/2.0 
    z1 = z1[1:].values.reshape((len(z1)-1,1))  
    B = np.append(-z1, np.ones_like(z1), axis = 1)
    Yn = x0[1:].reshape((len(x0)-1, 1))  
    [[a],[b]] = np.dot(np.dot(np.linalg.inv(np.dot(B.T, B)), B.T), Yn) 
    f = lambda k: (x0[0]-b/a)*np.exp(-a*(k-1))-(x0[0]-b/a)*np.exp(-a*(k-2)) 
    delta = np.abs(x0 - np.array([f(i) for i in range(1,len(x0)+1)]))  
    C = delta.std()/x0.std()
    P = 1.0*(np.abs(delta - delta.mean()) < 0.6745*x0.std()).sum()/len(x0)
    print( str(i) + ": C = " + str(C) + " , P = " +  str(P))
    return f, a, b, x0[0], C, P 


l=['x1','x2','x9'] #根据Lasso分析结果，选择变量
data_1=X[l].copy()
data_1.index=range(2004,2016)
data_1.loc[2016]=None  
data_1.loc[2017]=None

for i in l:  #预测每个变量2016、2017预测值
    f=GM11(data_1[i][:-2].values,i)[0]
    data_1[i][2016]=f(len(data_1)-1)
    data_1[i][2017]=f(len(data_1))
    data_1[i]=data_1[i].round(2)

print(data_1.tail(2))

#step5:SVR模型
result = data_1.copy()
result['real'] = y
result['predict'] = None #记录结果

from sklearn import svm
from sklearn.metrics import r2_score
svm_model = svm.SVR(kernel='linear',max_iter=10)
svm_model.fit(data_1.iloc[0:12, :], y)
predict1 = svm_model.predict(data_1.iloc[0:12, :])
r2_score = r2_score(y,predict1)


result.iloc[0:12,4]=predict1

from sklearn.metrics import r2_score
print('R2 = ' + str(r2_score(y,predict1) ))
plt.scatter(y.index, y, color='darkorange', label='data')
plt.plot(y.index, predict1, color='navy', lw=2, label='Linear model')
plt.xlabel('data')
plt.ylabel('target')
plt.title('Support Vector Regression')
plt.legend()
plt.show()

predict2 = svm_model.predict(data_1.iloc[12:14, :])
result.iloc[12:14,4]=predict2
print('最终结果：')
print(result)