Code Generation

Generate C/C++ code and MEX functions for Statistics and Machine Learning Toolbox™ functions

MATLAB® Coder™ generates readable and portable C and C++ code from Statistics and Machine Learning Toolbox functions that support code generation. For example, you can classify new observations on hardware devices that cannot run MATLAB by deploying a trained support vector machine (SVM) classification model to the device using code generation.

You can generate C/C++ code for the Statistics and Machine Learning Toolbox functions in several ways.

  • Code generation for the object function (predict, random, knnsearch, or rangesearch) of a machine learning model — Use saveLearnerForCoder, loadLearnerForCoder, and codegen. Save a trained model by using saveLearnerForCoder. Define an entry-point function that loads the saved model by using loadLearnerForCoder and calls the object function. Then use codegen to generate code for the entry-point function.

  • Code generation for the predict and update functions of a tree model, an SVM model, a linear model, or a multiclass error-correcting output codes (ECOC) classification model using SVM or linear binary learners — Create a coder configurer by using learnerCoderConfigurer and then generate code by using generateCode. You can update model parameters in the generated C/C++ code without having to regenerate the code.

  • Other functions that support code generation — Use codegen. Define an entry-point function that calls the function that supports code generation. Then generate C/C++ code for the entry-point function by using codegen.

You can also generate fixed-point C/C++ code for the prediction of an SVM classification model or an SVM regression model. This type of code generation requires Fixed-Point Designer™.

To learn about code generation, see Introduction to Code Generation.

Functions

expand all

saveLearnerForCoderSave model object in file for code generation
loadLearnerForCoderReconstruct model object from saved model for code generation
generateLearnerDataTypeFcnGenerate function that defines data types for fixed-point code generation

Create Coder Configurer Object

learnerCoderConfigurerCreate coder configurer of machine learning model

Work with Coder Configurer Object

generateCodeGenerate C/C++ code using coder configurer
generateFilesGenerate MATLAB files for code generation using coder configurer
validatedUpdateInputsValidate and extract machine learning model parameters to update
updateUpdate model parameters for code generation

Objects

expand all

ClassificationTreeCoderConfigurerCoder configurer of binary decision tree model for multiclass classification
ClassificationSVMCoderConfigurerCoder configurer for support vector machine (SVM) for one-class and binary classification
ClassificationLinearCoderConfigurerCoder configurer for linear binary classification of high-dimensional data
ClassificationECOCCoderConfigurerCoder configurer for multiclass model using binary learners
RegressionTreeCoderConfigurerCoder configurer of binary decision tree model for regression
RegressionSVMCoderConfigurerCoder configurer for support vector machine (SVM) regression model
RegressionLinearCoderConfigurerCoder configurer for linear regression model with high-dimensional data

Topics

Code-Generation-Enabled Functions

Code Generation Support, Usage Notes, and Limitations

View code generation usage notes, limitations, and the list of code-generation-enabled Statistics and Machine Learning Toolbox functions.

Code Generation Workflows

Introduction to Code Generation

Learn how to generate C/C++ code for Statistics and Machine Learning Toolbox functions.

General Code Generation Workflow

Generate code for Statistics and Machine Learning Toolbox functions that do not use machine learning model objects.

Code Generation for Prediction of Machine Learning Model at Command Line

Generate code for the prediction of a classification or regression model at the command line.

Code Generation for Prediction of Machine Learning Model Using MATLAB Coder App

Generate code for the prediction of a classification or regression model by using the MATLAB Coder app.

Code Generation for Prediction and Update Using Coder Configurer

Generate code for the prediction of a model using a coder configurer, and update model parameters in the generated code.

Code Generation and Classification Learner App

Train a classification model using the Classification Learner app, and generate C/C++ code for prediction.

Code Generation for Nearest Neighbor Searcher

Generate code for finding nearest neighbors using a nearest neighbor searcher model.

Specify Variable-Size Arguments for Code Generation

Generate code that accepts input arguments whose size might change at run time.

Train SVM Classifier with Categorical Predictors and Generate C/C++ Code

Convert categorical predictors to numeric dummy variables before fitting an SVM classifier and generating code.

Fixed-Point Code Generation for Prediction of SVM

Generate fixed-point code for the prediction of an SVM classification or regression model.

Code Generation for Probability Distribution Objects

Generate code that fits a probability distribution object to sample data and evaluates the fitted distribution object.

Code Generation Applications

Predict Class Labels Using MATLAB Function Block

Generate code from a Simulink® model that classifies data using an SVM model.

System Objects for Classification and Code Generation

Generate code from a System object™ for making predictions using a trained classification model, and use the System object in a Simulink model.

Predict Class Labels Using Stateflow

Generate code from a Stateflow® model that classifies data using a discriminant analysis classifier.

Featured Examples