Introduction

The GRAPE-library contains a large variety of methods for visualizing data given on hierachical grids. On the one hand there are methods for analysing the grid itself on the other hand functions defined on the grid can be presented in different ways. However, not all possible methods and algorithms for extracting and visualizing data can be included in advance. So in accordance with the open and extensible concept of GRAPE the user can write his own display methods. How to do this and which tools exist in GRAPE to facilitate this is the scope of the following sections.

In general any display method will have the following structure:

• Preparation, includes getting the object to be displayed, getting a graphic device, initializing interactive elements with which parameters of the display method can be altered and possibly getting additional tools as for example the colorbar.
• Data extraction, by means of the interface functions of the Hmesh - first-macro, next-macro, etc. - and the access function of the F_hdata-structure the necessary data is extracted. As this process is very similar in many display methods there are powerfull tools for the most common extraction requirements.
• Rendering, finally using the functionality of the graphic device the extracted data are rendered. This part of the display method is almost always specific for it and thus it is the part where except for the routines provided in the graphic device no other tools are available.

In the following each of the steps mentioned above is discussed in detail. Once the user has written a display-method it must be added to the appropriate class as described in [38]. In order to be able to choose the display method in the manager a naming convention must be complied with.

Most examples in this section are based on the class HMesh2d. However, the corresponding class HMesh3d is similar and most of the examples work also in the 3D-case.