scikit-learn

什么是scikit-learn?

Python的机器学习库，用于数据挖掘和数据分析的简单而有效的工具。

scikit-learn 提供了什么主要功能？

1. 回归
   -Lasso
   -Ridge
   -ElasticNet
2. 分类
   -SVM
   -RandomForestClassifier
   -NearestNeighbors

数据格式

from sklearn.linear_model import Lasso
model_Lasso = Lasso(alpha=0.1)
X = [[-1,-1], [0,0], [1,1]]
y = [-1,0,1]
model_Lasso.fit(X,y)

from sklearn.svm import SVC
import numpy as np
X = np.array([[-3,-2],[-4,-5],[3,4],[4,5]])
y = np.array([1, 1, 2, 2])
model_SVC = SVC()
model_SVC.fit(X, y)

常用代码

Rolling 计算Regression

from sklearn import linear_model
from sklearn.metrics import r2_score
reg = linear_model.Lasso(alpha=1)
result = result.dropna()
# print result
y = result.HS300_close.values
target = result[["ADP_MAS", "HLP_MAS", "MAP_MAS", 'VOL_MAS']]
data = map(lambda row: list(row[1]), target.iterrows())
residual=[]
for i in range(0, len(data)-100, 1):
    X = data[i:100+i]
#     print len(X)
    YY = y[i:100+i]
    reg.fit(X, YY)
    print reg.score(X, YY)
    rsd = YY - reg.predict(X)
    residual.append(rsd[-1])
res = pd.Series(residual, index=result.index[100:])

官方文档： http://scikit-learn.org/