Main Content

Execute MATLAB Functions from .NET

When you call a MATLAB® function from .NET, you can pass .NET variables to the function. If the MATLAB function returns an output variable, the engine converts it from MATLAB to .NET.

For information on how to set up and build .NET engine programs, see Test Your .NET Development Environment.

Pass Variables from .NET to MATLAB

In this example, you create a 1-D double array using the MATLAB linspace function and reshape the elements using reshape. For information about converting data from .NET to MATLAB, see Pass .NET Data Types to MATLAB Functions.

The MATLAB code for this example is:

A = linspace(-5.0,5.0);
sz = [25,4];
B = reshape(A,sz);

The C# code for this example is:

using MathWorks.MATLAB.Engine;
using MathWorks.MATLAB.Exceptions;
using MathWorks.MATLAB.Types;
using System;

namespace MathWorks.MATLAB.Engine.ConsoleExamples {
    public class Program {
        public static void Main(string[] args) {
            Console.Write("Starting MATLAB... ");
            using (dynamic matlab = MATLABEngine.StartMATLAB()) {
                Console.WriteLine("done.");
                double[] A = matlab.linspace(-5.0,5.0);
                int[] sz = new int[] {25,4};
                double[,] B = matlab.reshape(A,sz);
            }
            // Call when you no longer need MATLAB Engine in your application.
            MATLABEngine.TerminateEngineClient();
        }
    }
}

Pass Variables from MATLAB to .NET

The MATLAB magic function returns a 2-D matrix of type double. The engine converts the returned MATLAB 2-D double matrix to a .NET variable dbls declared as double[,]. You can then use dbls in your .NET program. For information about converting data from MATLAB to .NET, see Handle MATLAB Data in .NET Applications.

The C# code for this example is:

using MathWorks.MATLAB.Engine;
using MathWorks.MATLAB.Exceptions;
using MathWorks.MATLAB.Types;
using System;

namespace MathWorks.MATLAB.Engine.ConsoleExamples {
    public class Program {
        public static void Main(string[] args) {
            Console.Write("Starting MATLAB... ");
            using (dynamic matlab = MATLABEngine.StartMATLAB()) {
                Console.WriteLine("done.");
                double[,] dbls = matlab.magic(3.0);
                matlab.disp(new RunOptions(nargout: 0), dbls);
            }
            // Call when you no longer need MATLAB Engine in your application.
            MATLABEngine.TerminateEngineClient();
        }
    }
}

In this code, variable dbls is defined as a 2-D matrix of double. Alternative size definitions are:

double[,] dbls = matlab.magic(3.0);  // 3x3
double[,,] dbls = matlab.magic(3.0); // 3x3x1
Array dbls = matlab.magic(3.0);      // 3x3

This code shows size definitions for representing number 3.0:

double x = matlab.eval(" 3.0 ");       // double scalar
double[] y = matlab.eval(" 3.0 ");     // double array, length 1
double[,,,] z = matlab.eval(" 3.0 ");  // 1x1x1x1
Array q = matlab.eval(" 3.0 ");        // 1x1

Convert to Common Language Runtime Type

You can use the .NET IConvertible interface to convert the value of a reference or value type to a common language runtime type with an equivalent value.

Before R2024a: C# Code

C# Code (since R2024a)

MATLABArray retval = matlab.eval("int32(42)");
dynamic retvalDynamic = retval;
double d = retvalDynamic;

MATLABArray retval = matlab.eval("int32(42)");
double d = Convert.ToDouble(retval);

Widening conversion occurs when a value of one type is converted to another type that is of equal or greater size. The .NET type does not need to exactly match the MATLAB type, if the conversion is lossless.

double d = matlab.eval("int32(42)");
float f = matlab.eval("uint8(42)");

For conversion information, see MATLAB Numeric Types in .NET.

See Also

Related Topics