Programming Considerations for Code Generation

MATLAB^® programming considerations for code generation; behavior of generated code

Generating code from MATLAB for desktop and embedded systems enables you to perform your software design, implementation, testing, and deployment within the MATLAB environment. Due to differences in typing, array sizing, memory and other aspects of MATLAB and C/C++ languages, follow certain guidelines while designing your code in MATLAB for which you intend to generate C/C++ code. There are certain MATLAB language features that code generation does not support. Certain code generation optimizations can cause the generated code to behave differently than the source code. The code generation report and the MATLAB Coder™ app display the potential differences between generated code and MATLAB code.

To generate C/C++ or MEX code from MATLAB algorithms, you must have MATLAB Coder and a C/C++ compiler installed.

Topics

When to Generate Code from MATLAB Algorithms
Decide whether to generate code for your application.
MATLAB Code Design Considerations for Code Generation
When writing MATLAB for code generation, consider design constraints.
Differences Between Generated Code and MATLAB Code
Generated C/C++ code can behave differently than original MATLAB source code.
Potential Differences Reporting
The code generator identifies potential behavior differences between the generated code and the original MATLAB code.
Potential Differences Messages
Review potential difference messages.
MATLAB Language Features Supported for C/C++ Code Generation
Use the MATLAB language features and functions that code generation supports.