PML - General syntax conventions

General syntax conventions

Variable names are case sensitive and cannot contain special characters such as a period “.”. They can contain an underscore “_”, but if they do they are not compatible with S-PLUS syntax. The first character of a variable name cannot be an underscore “_”.

Column names are case sensitive and can contain special characters. However, if a column name contains a blank space, the data must be given in CSV format, and a special argument, /csv, must be given to the engine.

Column values (text or numeric) cannot contain a comma (‘,’).

Line boundaries are not significant. Statements can span multiple lines, except for comments. Characters that denote comments include.

# comment… end-of-line (S-PLUS convention)
/* comment… */ (multi-line, non-nesting, C convention)
// comment… end-of-line (C++ convention)

Block delimiters: { …} (curly brackets, S-PLUS convention)

Statement delimiter: An optional semicolon (S-PLUS convention)

Sub-statement delimiter: An optional comma

Assignment operators:

“=” sign (S-PLUS convention)
“<-” (S-PLUS convention)

Declaration of variables: Variable types are double precision so scoping is not needed (S-PLUS convention). Variables are of two types:

Declared variables are introduced by a declaration, such as deriv or real. These can be changed at points in time, such as in sequence statements.

Functional variables are introduced by being assigned at the top level of a model, such as C=A1/V1. They are regarded as being computed “all of the time.”

Model member reference: Models inherently act as structure. “$” is the model component reference operator (S-PLUS convention)

Real numbers: Although the Phoenix Modeling Language uses the real variable for designating real numbers, double is also acceptable.