时间序列实验指南new

《时间序列实验指南new》由会员上传分享，免费在线阅读，更多相关内容在行业资料-天天文库。

1、实验指南广东商学院统计系数量经济教研室编制二00一年六月二十八日47目录实验一分析太阳黑子数序列··········································3实验二模拟AR模型·················································4实验三模拟MA模型和ARMA模型····································6实验四分析化工生产量数据··········································8实验五模拟ARIMA模型和季节ARIMA模型··············

2、·············10实验六分析美国国民生产总值的季度数据·····························13实验七分析国际航线月度旅客总数数据·······························16实验八干预模型的建模·············································19实验九传递函数模型的建模47·········································22实验十回归与时序相结合的建模·····································25太阳黑子年度数据·····

3、·············································28美国国民收入数据··················································29化工生产过程的产量数据············································30国际航线月度旅客数据··············································30洛杉矶臭氧每小时读数的月平均值数据································31煤气炉数据················

4、········································35芝加哥某食品公司大众食品周销售数据································37牙膏市场占有率周数据··············································39某公司汽车生产数据················································4447加拿大山猫数据····················································44实验一分析太阳黑子数序列一、实验目的：了解

5、时间序列分析的基本步骤，熟悉SAS/ETS软件使用方法。二、实验内容：分析太阳黑子数序列。三、实验要求：了解时间序列分析的基本步骤，注意各种语句的输出结果。四、实验时间：2小时。五、实验软件：SAS系统。六、实验步骤1、开机进入SAS系统。2、创建名为exp1的SAS数据集，即在窗中输入下列语句：dataexp1;inputa1@@;year=intnx(‘year’,’1jan1742’d,_n_-1);formatyearyear4.;cards;输入太阳黑子数序列(见附表)run;3、保存此步骤中的程序，供以后分析使用（只需按工具条上的保存按钮然后填写完提问后就可以把这段程序

6、保存下来即可）。4、绘数据与时间的关系图，初步识别序列，输入下列程序：procgplotdata=exp1;symboli=splinev=starh=2c=green;plota1*year;run;472、提交程序，在graph窗口中观察序列，可以看出此序列是均值平稳序列。3、识别模型，输入如下程序。procarimadata=exp1;identifyvar=a1nlag=24;run;4、提交程序，观察输出结果。初步识别序列为AR(3)模型。5、估计和诊断。输入如下程序：estimatep=3;run;6、提交程序，观察输出结果。假设通过了白噪声检验，且模型合理，则进行预测

7、。7、进行预测，输入如下程序：forecastlead=6interval=yearid=yearout=out;run;procprintdata=out;run;8、提交程序，观察输出结果。9、退出SAS系统，关闭计算机。实验二模拟AR模型一、实验目的：熟悉各种AR模型的样本自相关系数和偏相关系数的特点，为理论学习提供直观的印象。二、实验内容：随机模拟各种AR模型。三、实验要求：记录各AR模型的样本自相关系数和偏相关系数，观察各种序列图形，总结AR模型的样本自相关