Java Reference
In-Depth Information
this .features # features;
}
public String name() { return name; }
public String value(String featureName){
// iterates over the features to find the value of the
// given one
for ( int i # 0; i<features.length; !! i){
if (featureName.equals(features[i].name())){
return features[i].value();
}
}
return null ;
}
public void setValue(String featureName, String newValue){
for ( int i # 0; i<features.length; !! i){
if (featureName.equals(features[i].name())){
features[i].value(newValue);
return ;
}
}
}
}
The structure of the decision tree is described by means of the classes
Node and Arc . Class Node has a read-only attribute ( label ) and is involved in
the one-to-many association children. Since the number of objects is not
known a priori it is not possible to use an array. We can implement the
association applying the transformation shown in Figure 4.8. The trans-
formed association can be easily implemented with a reference in the class
One plus a reference in the Many class: in total one attribute per class. This
is a standard way of implementing such associations and can be coded into
an idiom. This transformation is made during the implementation phase
Node
Node
label : String
label : String
0 .. 1
firstArc
0 .. n
+children
Arc
Arc
next
0 .. 1
label : String
label : String
Figure 4.8 One-to-many association transformation
 
Search WWH ::




Custom Search