把python的statsmodels通过json接口api的形

很多企业应用都是基于java开发的，如果需要进行基于网页的数据分析挖掘，通过java不是很方便，加上本人对java不熟悉，但公司应用又要用java实现数据分析的功能，我就想到一个方法通过java调用python接口去实现。

在python中，statsmodels的ols返回的摘要如下

import statsmodels.api as sm

import json

x1 = [[4,67,662],[9,19,618],[6,49,372],[6,33,58],[1,18,153],[2,78,938],[3,15,627],[8,55,191],[2,47,812],[2,83,946],[2,4,895],[9,37,42],[0,1,595],[7,27,392],[5,22,836],[0,12,513],[2,41,601],[3,68,615],[2,23,649],[1,98,9],[9,40,32],[5,77,798],[1,10,903],[1,53,772],[7,20,716],[2,35,678],[5,52,258],[7,31,814],[2,30,577]]

y1 = [2857.0163,2547.5962,1647.6061,343.8966,668.2108,3990.0414,2559.0662,945.1439,3393.1068,4037.1068,3596.0458,297.5798,2383.6193,1663.8839,3420.5135,2088.0197,2531.2703,2670.7878,2669.8044,332.9981,266.718,3433.975,3644.3636,3249.3518,2938.0325,2821.3308,1198.4373,3363.5752,2402.6042]

x1 = sm.add_constant(x1)

model = sm.OLS(y1, x1)

rs = model.fit()

现在以statsmodels的OLS最少二乘法的回归模型为例子，模型需要的参数有两个，x和y

http://127.0.0.1:5000/apis_statsmodels_ols/?x=[[4,67,662],[9,19,618],[6,49,372],[6,33,58],[1,18,153],[2,78,938],[3,15,627],[8,55,191],[2,47,812],[2,83,946],[2,4,895],[9,37,42],[0,1,595],[7,27,392],[5,22,836],[0,12,513],[2,41,601],[3,68,615],[2,23,649],[1,98,9],[9,40,32],[5,77,798],[1,10,903],[1,53,772],[7,20,716],[2,35,678],[5,52,258],[7,31,814],[2,30,577]]&y=[2857.0163,2547.5962,1647.6061,343.8966,668.2108,3990.0414,2559.0662,945.1439,3393.1068,4037.1068,3596.0458,297.5798,2383.6193,1663.8839,3420.5135,2088.0197,2531.2703,2670.7878,2669.8044,332.9981,266.718,3433.975,3644.3636,3249.3518,2938.0325,2821.3308,1198.4373,3363.5752,2402.6042]

返回的是

详细的返回内容说明，请参考https://www.statsmodels.org/dev/generated/statsmodels.regression.linear_model.OLS.html#statsmodels.regression.linear_model.OLS

实现方法请参考我的github项目。https://github.com/stjnu/statsmodels_json

项目正在不断增加statsmodel的模型。