Graphing Data

Constructing graphs from data is an important part of the process of data analysis and presentation. Results displayed in graphs are generally more concise and often more illuminating; one may be able to detect patterns and relationships in data that are not readily apparent from examining tables of summary statistics.

Accordingly, EViews provides an easy-to-use, full-featured set of tools for the graphical display of information. With EViews, you can quickly and easily display graphs of data, customize those graphs, and output the results so that they may be incorporated into your presentations.

There are many aspects of the graphing of data in EViews. This chapter describes the basics of graphing data in series and groups of series using the View/Graph... menu item. (Most of the graphs in this chapter may also be generated from vectors and matrices, but for brevity, we will speak mostly of series and groups).

Three types of graphs are described in this chapter:

• Observation graphs which show the data for each observation in the series or group. A line plot of the observations in a series or a scatterplot of observations for pairs of series in a group are examples of observation graphs.

• Analytical graphs where we display results obtained from analysis of the series or group data. You might, for example, show a histogram or a boxplot computed from the original data.

• Auxiliary graphs are analytical graphs that are not meant to stand alone, but rather are to be added to existing observation graphs. For example, we may display the linear regression or kernel fit line on top of a scatterplot. Strictly speaking, auxiliary graphs are not a graph type, but rather a modification of an existing observation plot.

A fourth class of graphs, categorical graphs, consists of observation or analytical graphs formed using data divided into categories defined by factor variables. Categorical graphs are described in “Categorical Graphs”.

We do not consider here the specialized series and group routines that produce graphical output. For example, views of an equation produce graphs of the equation forecasts and residuals. Similarly, views of a model object that show graphs of simulation results, and the views of a state space object that show estimated states or signals. These graphs are described in the context of the specific views and procedures.

The remainder of this chapter is structured as follows. The first section offers a quick overview of the process of constructing a series or group graph view. The next two sections describe the process of constructing graphs from series and groups in somewhat greater depth. Next, we describe a handful of the most commonly performed graph view customizations. The final section provides detail on individual graph types.