Graphics Reference
In-Depth Information
11
Appendix
The appendix contains a hodgepodge of material that did not fit well in any of the previ-
ous chapters. We begin by illustrating some of the other kinds of graphs Stata can produce
that were not covered in this topic and how to use the options illustrated in this topic to
make them. Next, we look at how to save graphs, redisplay graphs, and combine multiple
graphs into a single graph. This is followed by a section with more realistic examples that
require a combination of multiple options or data manipulation to create the graph. We
review some common mistakes in writing graph commands and showing how to fix them,
followed by a brief look at creating custom schemes. This chapter and the topic conclude
by describing the online supplements to the topic and how to get them.
11.1
Overview of statistical graph commands, stat graphs
This section illustrates some of the Stata commands for producing specialized statistical
graphs. Unlike other sections of this topic, this section merely illustrates these kinds of
graphs but does not further explain the syntax of the commands used to create them. The
graphs are illustrated on the following six pages, with multiple graphs on each page. The
title of each graph is the name of the Stata command that produced the graph. We can use
the
help
command to find out more about that command or look up more information in
the appropriate Stata manual. The figures are described below.
•
Figure 11.1 illustrates a number of graphs used to examine the univariate distribution
of variables.
•
Figure 11.2 illustrates the
gladder
and
qladder
commands, which show the dis-
tribution of a variable according to the
ladder of powers
to help visually identify
transformations for achieving normality.
•
Figure 11.3 shows a number of graphs that can be used to assess how your data meets
the assumptions of linear regression.
•
Figure 11.4 shows some plots that help to illustrate the results of a survival analysis.
•
Figure 11.5 shows a number of different plots used to understand the nature of time-
series data and to select among different time-series models.
•
Figure 11.6 shows plots associated with Receiver Operating Characteristic (
ROC
) anal-
yses, which can also be used with logistic regression analysis.
345
The electronic form of this topic is solely for direct use at UCLA and only by faculty, students, and staff of UCLA.
