EViews 5.1: Überblick

When you ar ready to begin estimation, EViews allows you to choose from a full set of single

equation estimators, including ordinary and nonlinear least squares (multiple regression), weighted

least squares, two-stage least squares and Generalized Method of Moments (GMM). For time

series analysis, EViews estimates ARMA, ARMAX and a variety of ARCH specifications.  Limited

dependent variable models for binary, ordered, censored (Tobit), truncated, and count data are

also available.  Longitudinal data are handled by a wide range of pooled time series and panel

estimators including modern Dynamic Panel Data (DPD) techniques.

EViews also offers powerful tools for analyzing systems of equations. You can use EViews

to estimate linear and nonlinear systems by least squares, two-stage least squares, seemingly

unrelated regression, three-stage least squares, GMM, and FIML. Additional support is pro­vided

for VAR and VEC estimation, state space estimation and Kalman filtering.

   Explore the time series properties of your data.

For custom analysis, EViews’ easy-to-use likelihood object permits estimation of user­specified

maximum likelihood models. You simply provide standard EViews expressions to describe

the log likelihood contributions, set coefficient starting values, and EViews will do the rest.

Once you estimate your equation or system, EViews makes it easy to perform specification

and diagnostic tests. You can evaluate coeffi­cient restrictions, perform residual diagnostics,

examine your specification for stability, and conduct a variety of estimator specific speci­fication

tests.  Even if your favorite test is not supported directly, easy access to coefficient values, residuals,

gradients, and other estima­tion results makes it easy to construct your own diagnostics.

Forecasting and Simulation

EViews frees you from the complexities of forecasting, allowing you to concentrate on the substance

of your problem.

You can generate static or dynamic forecasts from an estimated equation with just a few mouse clicks.

Forecast standard errors are calculated automatically and a graph of the 95 percent confidence interval and

a variety of forecast evaluation statistics are  also available.  If the dependent variable of your equation has

been transformed, for example by taking logarithms or differencing, you can easily generate forecasts for either

the trans­formed dependent variable or the underlying series.

Multiple equation forecasting is handled by the EViews model object.  Comparable in function to the best

stand-alone programs for simulation of simultaneous equations, the EViews model object provides the tools

demanded by professional model builders for managing, analyzing and solving large systems of equations.

Models offer a variety of solution methods, including stochastic simulation and forward solution for model

consistent expectations, as well as tools for managing alternative solution scenarios and user-specified add

factors.  Models also allow you to trace the dependency structure of the endogenous variables, to solve simple

control problems, and to generate custom tables and graphs that compare solution results under alternative

assumptions.

Models may be tightly integrated with EViews estimation objects. While you can specify the equations of a

model by typing in simple algebraic expressions, it is also possible to link previously estimated equation,

system or VAR objects to a model, so that the model automatically updates whenever its underly­ing equations

are re-estimated.

Sophisticated Data Managementore

From its extensive library of mathematical, statistical, date, string, and time series opera­tors and functions,

to  comprehensive support for numeric, character, and date data, EViews offers the data handling features

you’ve come to expect from modern statistical software.

EViews goes beyond the conventional, allow­ing you to effortlessly manage complex data structures involving

regular or irregular dated data (for example, trading day), cross-section data with observation identifiers, and

panel data. EViews’ state-of-the-art expression handling and linking features even allow for dynamic formula

evaluation, match merging, and automatic frequency conversion.

Exchanging data with other programs is easy, since EViews reads and writes over 20 popular data formats.

Simply drag-and-drop your foreign file onto EViews and your data will automatically appear in an EViews

workfile.  Or share your EViews data with others by writing your workfile to Microsoft Excel or Access. You

can even use the EViews Enterprise Edition to read data directly from a relational database using an ODBC

query statement or by opening an ODBC dsn or query file.

Presentation Quality Output

EViews 5.1 supports a wide range of basic graph types including line graphs, bar graphs, filled area graphs,

pie charts, scatter diagrams, mixed line-bar graphs, high-low graphs, scatterplots, and boxplots.  Any number

of graphs can be combined in a single graph for presentation. Various options give you control over line types,

symbols, color, frame and border characteristics, headings, shading, and scaling, including logarithmic scaling

and dual scale graphs.Legends are automatically created and you can add labels in any scalable Windows

font anywhere on your graph.

   Create custom graphs for inclusion in other applications.

Customizing a graph is as simple as modifying or moving graphic elements on the screen. Everything from

aspect ratios, to line and symbol characteristics, to axes scaling and labeling is right at your fingertips. Want

to change the font or other characteristics of a legend or a text label? Just click on an element of the graph,

and your choices are presented in an easy to understand dialog.  You can even use a customized graph

template to modify all of your graphsettings at once.

You can quickly incorporate customized graphs into other applications using copy-and-paste or by writing the

graph to a Windows metafile or PostScript file.