The plots in the GTL
are classified in different ways, depending on the context of the
discussion.
Within layout blocks, plots are often
classified according to graphical dimension: whether they are projected
in two or three visual dimensions. Thus, plots in the GTL are often
referred to as 2-D or 3-D plots, based on their graphical dimensions,
not their data dimensions.
Relative
to their input data, plots are classified according to the statements
that calculate summary statistics from raw input data, and those that
use calculated statistics as input parameters on the plot statement.
Thus, many GTL plot statements have two versions: BARCHART and BARCHARTPARM,
HISTOGRAM and HISTOGRAMPARM, and so on. The main distinction between
such plots is the nature of the input data that they accept:
-
The “non-parm” version
(for example, BARCHART) computes its values from raw, unsummarized
data. For example, a BARCHART computes the summary values it needs
for the bars in the chart. Such plots are often referred to as “computed
plots.”
-
The “parm” version
(for example, BARCHARTPARM) does not summarize or compute values from
the input data but instead simply renders the input data it is given.
Thus, the input data must be pre-summarized, perhaps by a SAS procedure.
The “parm” version of plots, often referred to as “parameterized
plots,” produce the same result as the non-parm version. However,
they don’t perform the calculations or data summarizations
needed to achieve the result.
Key Concepts for Using Plots discusses general concepts that apply across plot types.
For detailed information about a particular plot, see the chapter
for that plot.