Function Definition
R2026bThere are special considerations when you create MATLAB® functions that are intended for code generation. These include certain
restrictions when you use varargin and
varargout, recursive functions, anonymous functions, and
nested functions. To generate code for MATLAB recursive functions, the code generator uses either compile-time
recursion or run-time recursion. To produce efficient C/C++ code, the code generator
sometimes creates multiple versions of a function, referred to as function
specializations. In certain cases, you can prevent the code generator from creating
function specializations by using coder.ignoreSize and coder.ignoreConst. While generating a MEX function, you can bypass
code generation for a specific function in your MATLAB code and instead use the MATLAB engine to execute the call. To do this, use the coder.extrinsic construct.
Functions
%#codegen | Pragma that Indicates a MATLAB function is intended for code generation |
coder.extrinsic | Declare extrinsic function |
coder.ignoreConst | Prevent use of constant value of expression for function specializations |
coder.ignoreSize | Prevent code generator from creating function specializations for constant-size expressions |
coder.mustBeComplex | Validate that value lies on the complex plane (Since R2023b) |
coder.mustBeConst | Validate that value is a compile-time constant (Since R2023b) |
coder.noImplicitExpansionInFunction | Disable implicit expansion within the specified function in the generated code |
coder.sameSizeBinaryOp | Apply element-wise binary operations without implicit expansion |
coder.specifyAsGPU | Specify that value is a GPU input to an entry-point function for GPU code generation (Since R2023b) |
Topics
Implicit Expansion
- Effects of Implicit Expansion on Generated Code
Implicit expansion in the generated code can affect output sizes, code size, and performance. - Control Implicit Expansion in Generated Code
Disable implicit expansion globally, for specific functions, or for specific operations.
Extrinsic Functions
- Using Extrinsic Functions
Use MATLAB to execute a function call in the generated code. - Working With mxArray Outputs in Generated Code
Handle themxArrayobjects returned by extrinsic functions. - Use Extrinsic Functions to Calibrate Data for Standalone Code Generation
Use MATLAB to execute function calls in the generated code.
Function Argument Validation
- Generate Code for arguments Block That Validates Input and Output Arguments
Generate code for MATLAB code that constrains class, size, and other aspects of function input and output values.
Recursive, Anonymous, and Nested Functions
- Code Generation for Recursive Functions
Use recursive functions in MATLAB code that is intended for code generation. - Code Generation for Nested and Anonymous MATLAB Functions
Restrictions apply when you use nested or anonymous functions in MATLAB code intended for code generation.
Troubleshooting
Resolve Error: Nonconstant Index into varargin or varargout
Troubleshoot error when the code generator cannot determine the value of an index
into varargin or varargout inside a
for-loop.
Troubleshoot Replicated Functions in Generated Code
Troubleshoot duplicate functions in the generated code.
Resolve Error: Output Variable Must Be Assigned Before Run-Time Recursive Call
Troubleshoot error when the MATLAB code does not assign a value to an output variable before a recursive call.
Resolve Error: Compile-Time Recursion Limit Reached
Troubleshoot error when recursion during code generation exceeds the limit.
Resolve Error: Arrays Have Incompatible Sizes
Troubleshoot size incompatibility errors that occur during code generation.
Resolve Issue: coder.inline("never") and coder.nonInlineCall Do Not Prevent Function Inlining
Troubleshoot instances of coder.inline('never') not
preventing inlining.
Resolve Error: Variables Must Be Fully Defined Before Use
Troubleshoot code generation error when variables are not defined before use.
Resolve Error: Cannot Determine the Exact Number of Iterations for a Loop
Troubleshoot error when code generator is unable to determine the number of iterations in a loop.
