Eviews操作完整操作指引 - 图文

1. EVIEWS基础···························································································· 3 1.1. 1.2. 1.3. 1.4. 1.5.

EVIEWS简介 ························································································ 3 EVIEWS的启动、主界面和退出 ································································ 3 EVIEWS的操作方式 ··············································································· 5 EVIEWS应用入门 ·················································································· 6 EVIEWS常用的数据操作 ········································································ 15

2. 一元线性回归模型 ····················································································· 24 2.1. 2.2. 2.3.

用普通最小二乘估计法建立一元线性回归模型 ··········································· 24 模型的预测 ························································································ 30 结构稳定性的CHOW检验 ······································································ 34

3. 多元线性回归 ·························································································· 39 3.1. 3.2.

用OLS建立多元线性回归模型 ······························································· 39 函数形式误设的RESET检验 ································································· 45

4. 非线性回归 ····························································································· 48 4.1. 4.2. 4.3. 4.4. 4.5. 4.6.

用直接代换法对含有幂函数的非线性模型的估计 ········································ 48 用间接代换法对含有对数函数的非线性模型的估计 ····································· 50 用间接代换法对CD函数的非线性模型的估计 ··········································· 53 NLS对可线性化的非线性模型的估计 ······················································· 55 NLS对不可线性化的非线性模型的估计 ···················································· 58 二元选择模型 ····················································································· 62

5. 异方差 ···································································································· 68 5.1. 5.2. 5.3.

异方差的戈得菲尔德——匡特检验 ·························································· 68 异方差的WHITE检验 ·········································································· 72 异方差的处理 ····················································································· 75

6. 自相关 ···································································································· 79 6.1. 6.2.

自相关的判别 ····················································································· 79 自相关的修正 ····················································································· 83

7. 多重共线性 ····························································································· 87 7.1. 7.2.

多重共线性的检验 ··············································································· 87 多重共线性的处理 ··············································································· 92

8. 虚拟变量································································································· 94 8.1. 8.2. 8.3.

虚拟自变量的应用 ··············································································· 94 虚拟变量的交互作用 ············································································ 99 二值因变量：线性概率模型 ································································· 101

9. 滞后变量模型 ························································································ 105 9.1. 9.2.

自回归分布滞后模型的估计 ································································· 105 多项式分布滞后模型的参数估计 ··························································· 110

10. 联立方程模型 ······················································································· 115 10.1. 联立方程模型的单方程估计方法 ··························································· 115 10.2. 联立方程模型的系统估计方法 ······························································ 119

2

1. Eviews基础

1.1. Eviews简介

Eviews：Econometric Views（经济计量视图），是美国QMS公司（Quantitative Micro Software Co.，网址为http://www.eviews.com）开发的运行于Windows环境下的经济计量分析软件。Eviews是应用较为广泛的经济计量分析软件——MicroTSP的Windows版本，它引入了全新的面向对象概念，通过操作对象实现各种计量分析功能。

Eviews软件功能很强，能够处理以时间序列为主的多种类型数据，进行包括描述统计、回归分析、传统时间序列分析等基本数据分析以及建立条件异方差、向量自回归等复杂的计量经济模型。

1.2. Eviews的启动、主界面和退出 1.2.1. Eviews的启动

单击Windows的【开始】按钮，选择【程序】选项中的【Eviews 5】，单击其中的【Eviews5】；或者在相应目录下用鼠标双击所示：

标题栏 菜单栏 命令窗口 启动Eviews 5程序，进入主窗口。如图1.1

工作区 状态栏

图 1.1

3

1.2.2. Eviews的主界面

1.2.2.1. 标题栏

Eviews窗口的顶部是标题栏，标题栏左边是控制框；右边是控制按钮，有【最小化】、【最大化（或还原）】、【关闭】三个按钮。

1.2.2.2. 菜单栏

标题栏下面是菜单栏。菜单栏中排列着按照功能划分的9个主菜单选项，用鼠标单击任意选项会出现不同的下拉菜单，显示该部分的具体功能。9个主菜单选项提供的主要功能如下：

【File】 有关文件（工作文件、数据库、Eviews程序等）的常规操作，如文件的建立（New）、打开（Open）、保存（Save/Save As）、关闭（Close）、导入（Import）、导出（Export）、打印（Print）、运行程序（Run）等；选择下拉菜单中的Exit将退出Eviews软件。

【Edit】 通常情况下只提供复制功能（下拉菜单中只有Cut、Copy项被激活），应与粘贴（Paste）配合使用；对某些特定窗口，如查看模型估计结果的表达式时，可对窗口中的内容进行剪切（Cut）、删除（Delete）、查找（Find）、替换（Replace）等操作，选择Undo表示撤销上步操作。

【Objects】 提供关于对象的基本操作。包括建立新对象（New Objects）、从数据库获取/更新对象（Fetch/Update from DB）、重命名（Rename）、删除（Delete）。 【View】和【Procs】 二者的下拉菜单项目随当前窗口不同而改变，功能也随之变化，主要涉及变量的多种查看方式和运算过程。我们将在以后的实验中针对具体问题进行具体介绍。

【Quick】 下拉菜单主要提供一些简单常规用法的快速进入方式。如改变样本范围（Sample）、生成新序列（Generate Series）、显示对象（Show）、作图（Graph）、生成新组（Empty Group）以及序列和组的描述统计量、新建方程和VAR。

【Options】 系统参数设定选项。与一般应用软件相同，Eviews运行过程中的各种状态，如窗口的显示模式、字体、图像、电子表格等都有默认的格式，用户可以根据需要选择Options下拉菜单中的项目对一些默认格式进行修改。

【Windows】 提供多种在打开窗口种进行切换的方式，以及关闭所有对象（Close All Objects）或关闭所有窗口（Close All）。 【Help】 Eviews的帮助选项。选择Eviews Help Topics按照索引或目录方式在所有帮助信息种查找所需项目。下拉菜单还提供分类查询方式，包括对象（Object）、命令（Command）、函数（Function）、矩阵与字符串（Matrix&String）、程序（Programming）等五个方面。

4

1.2.2.3. 命令窗口

菜单栏下面是命令窗口（Command Windows），窗口内闪烁的“︱”是光标。用户可在光标位置用键盘输入各种Eviews命名，并按回车键执行该命令。

1.2.2.4. 工作区窗口

命令窗口下面是Eviews的工作区窗口。操作过程中打开的各子窗口将在工作区内显示。

1.2.2.5. 状态栏

Eviews主窗口的底部是状态栏，从左到右分别为：信息框、路径框、当前数据库框和当前工作文件框。

1.2.3. Eviews的退出

选择【File】项下的【Exit】将退出Eviews程序，或者直接点击标题栏上的【关闭】按钮如果工作文件没有保存，系统将提示用户保存文件。如图1.2所示

图 1.2

1.3. Eviews的操作方式

方式 说明 面向对象 5

面向过程