The Schematic Capture System (SCS) provides an integrated development environment (IDE) to build hierarchical neural models either by a top-down or bottom-up approach. SCS consists of the Schematic Editor, Icon Editor, NSLM Editor, Library Manager, and NSLM Viewer. SCS allows one to build a model graphically by connecting icons together into what we call a schematic. Each icon can then be decomposed further into a schematic of its own.
The Schematic Capture System (SCS) is an important component of the NSL system. SCS is primarily used to generate NSL models as shown in Figure 1.
The latest stable SCS distribution can be obtained from the SCS page. The installation guide describes how to run the GUI installer.
The Schematic Editor is responsible for building the structure of the model. It is also serves as the control window for the Schematic Capture System. From the Schematic Editor window we can start any of the other SCS tools, load a model or module into the Schematic Editor, build and simulate or debug a model, define a new library, or descend/ascend into a schematic. When selecting icons or modules, the Schematic Editor allows the user to pick which version of a module to use: the user can choose a floating version that can change at any time or a fixed version, which cannot change.
When opening other tools from the Schematic Editor it is import to note that each new tool pops up in its own window and we can have as many open as we would like. However, there is always one and only one Schematic Editor Window open at any time.
The Icon Editor allows the user to modify the graphical appearance of the individual icons (modules).
The NSLM Editor allows the user to add NSLM code to the code that SCS has generated. This is particularly important for “leaf” level modules since they contain most of the functionality of the module.
The Library Manager allows the user to access and create new libraries of models and modules within the file system.
The NSLM Viewer displays the code generated by SCS.
The walkthrough contains step-by-step instructions for creating the MaxSelector module of the MaxSelectorModel.