Methods to Initialize New Objects from a Class

Description

The arguments to function new to create an object from a particular class can be interpreted specially for that class, by the definition of a method for function initialize for the class. This documentation describes some existing methods, and also outlines how to write new ones.

Methods

signature(.Object = "ANY")

The default method for initialize takes either named or unnamed arguments. Argument names must be the names of slots in this class definition, and the corresponding arguments must be valid objects for the slot (that is, have the same class as specified for the slot, or some superclass of that class). If the object comes from a superclass, it is not coerced strictly, so normally it will retain its current class (specifically, as(object, Class, strict = FALSE)).

Unnamed arguments must be objects of this class, of one of its superclasses, or one of its subclasses (from the class, from a class this class extends, or from a class that extends this class). If the object is from a superclass, this normally defines some of the slots in the object. If the object is from a subclass, the new object is that argument, coerced to the current class.

Unnamed arguments are processed first, in the order they appear. Then named arguments are processed. Therefore, explicit values for slots always override any values inferred from superclass or subclass arguments.

signature(.Object = "traceable")

Objects of a class that extends traceable are used to implement debug tracing (see class traceable and trace).

The initialize method for these classes takes special arguments def, tracer, exit, at, print. The first of these is the object to use as the original definition (e.g., a function). The others correspond to the arguments to trace.

signature(.Object = "environment"), signature(.Object = ".environment")

The initialize method for environments takes a named list of objects to be used to initialize the environment. Subclasses of "environment" inherit an initialize method through ".environment", which has the additional effect of allocating a new environment. If you define your own method for such a subclass, be sure either to call the existing method via callNextMethod or allocate an environment in your method, since environments are references and are not duplicated automatically.

signature(.Object = "signature")

This is a method for internal use only. It takes an optional functionDef argument to provide a generic function with a signature slot to define the argument names. See Methods_Details for details.

Writing Initialization Methods

Initialization methods provide a general mechanism corresponding to generator functions in other languages.

The arguments to initialize are .Object and .... Nearly always, initialize is called from new, not directly. The .Object argument is then the prototype object from the class.

Two techniques are often appropriate for initialize methods: special argument names and callNextMethod.

You may want argument names that are more natural to your users than the (default) slot names. These will be the formal arguments to your method definition, in addition to .Object (always) and ... (optionally). For example, the method for class "traceable" documented above would be created by a call to setMethod of the form:

    setMethod("initialize", "traceable",
      function(.Object, def, tracer, exit, at, print) \dots
    )

In this example, no other arguments are meaningful, and the resulting method will throw an error if other names are supplied.

When your new class extends another class, you may want to call the initialize method for this superclass (either a special method or the default). For example, suppose you want to define a method for your class, with special argument x, but you also want users to be able to set slots specifically. If you want x to override the slot information, the beginning of your method definition might look something like this:

    function(.Object, x, ...) {
      Object <- callNextMethod(.Object, ...)
      if(!missing(x)) { # do something with x

You could also choose to have the inherited method override, by first interpreting x, and then calling the next method.