Main Content

matlabFunction

Convert symbolic expression to function handle or file

Description

example

ht = matlabFunction(f) converts the symbolic expression or function f to a MATLAB® function with handle ht. If there is an equivalent MATLAB function operating on the double data type for the symbolic expression or function, then the converted function can be used without Symbolic Math Toolbox™.

example

ht = matlabFunction(f1,...,fN) converts f1,...,fN to a MATLAB function with N outputs. The function handle is ht. Each element of f1,...,fN can be a symbolic expression, function, or a vector or matrix of symbolic expressions or functions.

example

ht = matlabFunction(___,Name,Value) specifies options using one or more name-value arguments in addition to any of the input argument combinations in the previous syntaxes.

For example, you can specify the File name-value argument to write the generated MATLAB function to a file. You can also specify the Vars name-value argument to generate a MATLAB function with input arguments that are a combination of scalar and vector variables.

Examples

collapse all

Convert the symbolic expression r to a MATLAB function with the handle ht. The converted function can be used without Symbolic Math Toolbox.

syms x y
r = sqrt(x^2 + y^2);
ht = matlabFunction(r)
ht =
  function_handle with value:
    @(x,y)sqrt(x.^2+y.^2)

Convert multiple symbolic expressions using comma-separated input.

ht = matlabFunction(r, r^2)
ht =
  function_handle with value:
    @(x,y)deal(sqrt(x.^2+y.^2),x.^2+y.^2)

Create a symbolic function and convert it to a MATLAB function with the handle ht.

syms x y
f(x,y) = x^3 + y^3;
ht = matlabFunction(f)
ht = 
  function_handle with value:
    @(x,y)x.^3+y.^3

Write the generated MATLAB function to a file by specifying the File name-value argument. Existing files are overwritten. When writing to a file, matlabFunction optimizes the code using intermediate variables named t0, t1, and so on. Include comments in the file by using the Comments option.

Write the MATLAB function generated from f to the file myfile.

syms x
f = x^2 + log(x^2);
matlabFunction(f,'File','myfile');
function f = myfile(x)
%MYFILE
%    F = MYFILE(X)

%    This function was generated by the Symbolic Math Toolbox version 8.4.
%    01-Sep-2019 00:00:00

t2 = x.^2;
f = t2+log(t2);

Include the comment Version: 1.1 in the file.

matlabFunction(f,'File','myfile','Comments','Version: 1.1');
function f = myfile(x)
...
%Version: 1.1
t2 = x.^2;
...

When converting a symbolic expression to a MATLAB function, you can specify the order of the input arguments of the resulting function. You can also specify that some input arguments are vectors instead of scalar variables.

Create a symbolic expression.

syms x y z
r = x + y/2 + z/3;

Convert r to a MATLAB function and write this function to the file myfile. By default, matlabFunction uses alphabetical order for the input arguments when converting symbolic expressions that contain only lowercase letters for the variable names. The generated input arguments are scalar variables x, y, and z.

matlabFunction(r,'File','myfile');
function r = myfile(x,y,z)
%MYFILE
%    R = MYFILE(X,Y,Z)
r = x+y./2.0+z./3.0;

Specify the Vars name-value argument as the vector [y z x] to modify the order of the input arguments for the generated MATLAB function. The generated input arguments are scalar variables y, z, and x.

matlabFunction(r,'File','myfile','Vars',[y z x]);
function r = myfile(y,z,x)
%MYFILE
%    R = MYFILE(Y,Z,X)
r = x+y./2.0+z./3.0;

Now, convert a symbolic expression v to a MATLAB function whose input arguments are a scalar and a vector. Specify the Vars name-value argument as the cell array {t,[x y z]}. The generated input arguments are a scalar variable t and a 1-by-3 vector variable in2.

syms x y z t
v = (x + y/2 + z/3)*exp(-t);
matlabFunction(v,'File','myfile','Vars',{t,[x y z]});
function v = myfile(t,in2)
%MYFILE
%    R = MYFILE(T,IN2)
x = in2(:,1);
y = in2(:,2);
z = in2(:,3);
v = exp(-t).*(x+y./2.0+z./3.0);

To generate a MATLAB function with input arguments that are vector variables, specify the Vars name-value argument as a cell array.

Create a symbolic expression that finds the dot product of two 1-by-3 vectors.

syms x y [1 3] real
f = dot(x,y);

Convert the expression f to a MATLAB function. Specify Vars as a cell array {x,y}. The generated input arguments are two 1-by-3 vector variables in1 and in2 that correspond to x and y, respectively.

matlabFunction(f,'File','myfile','Vars',{x,y});
function f = myfile(in1,in2)
%MYFILE
%    F = MYFILE(IN1,IN2)
x1 = in1(:,1);
x2 = in1(:,2);
x3 = in1(:,3);
y1 = in2(:,1);
y2 = in2(:,2);
y3 = in2(:,3);
f = x1.*y1+x2.*y2+x3.*y3;

Now create a symbolic function that is a function of four variables.

syms g(x,y,z,t)
g(x,y,z,t) = x^2 + y^2 + z^2 - t^2;

Convert the function g to a MATLAB function. To specify the generated input arguments as a 4-by-1 column vector from the input variables of g, specify Vars as a cell array that contains the 4-by-1 vector. You can use argnames to automatically obtain the input variables x, y, z, and t from g.

matlabFunction(g,'File','myfunction','Vars',{argnames(g).'});
function g = myfunction(in1)
%MYFUNCTION
%    G = MYFUNCTION(IN1)
x = in1(1,:);
y = in1(2,:);
z = in1(3,:);
t = in1(4,:);
g = -t.^2+x.^2+y.^2+z.^2;

When you convert a symbolic expression to a MATLAB function and write the resulting function to a file, matlabFunction optimizes the code by default. This approach can help simplify and speed up further computations that use the file. However, generating the optimized code from some symbolic expressions and functions can be time-consuming. Use Optimize to disable code optimization.

Create a symbolic expression.

syms x
r = x^2*(x^2 + 1);

Convert r to a MATLAB function and write the function to the file myfile. By default, matlabFunction creates a file containing the optimized code.

f =  matlabFunction(r,'File','myfile');
function r = myfile(x)
%MYFILE
%    R = MYFILE(X)
t2 = x.^2;
r = t2.*(t2+1.0);

Disable the code optimization by setting the value of Optimize to false.

f =  matlabFunction(r,'File','myfile','Optimize',false);
function r = myfile(x)
%MYFILE
%    R = MYFILE(X)
r = x.^2.*(x.^2+1.0);

When you convert a symbolic matrix to a MATLAB function, matlabFunction represents the symbolic matrix with a dense matrix by default. If most of the elements of the input symbolic matrix are zeros, the more efficient approach is to represent the symbolic matrix with a sparse matrix.

Create a 3-by-3 symbolic diagonal matrix.

syms x
A = diag(x*ones(1,3))
A =
[ x, 0, 0]
[ 0, x, 0]
[ 0, 0, x]

Convert A to a MATLAB function representing a numeric matrix, and write the result to the file myfile1. By default, the generated MATLAB function creates a dense numeric matrix specifying each element of the matrix, including all zero elements.

f1 = matlabFunction(A,'File','myfile1');
function A = myfile1(x)
%MYFILE1
%    A = MYFILE1(X)
A = reshape([x,0.0,0.0,0.0,x,0.0,0.0,0.0,x],[3,3]);

Convert A to a MATLAB function by setting Sparse to true. Now, the generated MATLAB function creates a sparse numeric matrix specifying only nonzero elements and assuming that all other elements are zeros.

f2 = matlabFunction(A,'File','myfile2','Sparse',true);
function A = myfile2(x)
%MYFILE2
%    A = MYFILE2(X)
A = sparse([1,2,3],[1,2,3],[x,x,x],3,3);

When converting a symbolic expression to a MATLAB function, you can specify the names of the output variables. Note that matlabFunction without the File argument (or with a file path specified as an empty character vector) creates a function handle and ignores the Outputs flag.

Create symbolic expressions r and q.

syms x y z
r = x^2 + y^2 + z^2;
q = x^2 - y^2 - z^2;

Convert r and q to a single MATLAB function and write the resulting function to a file myfile, which returns a vector of two elements, name1 and name2.

f = matlabFunction(r,q,'File','myfile', ...
                   'Outputs',{'name1','name2'});
function [name1,name2] = myfile(x,y,z)
%MYFILE
%    [NAME1,NAME2] = MYFILE(X,Y,Z)
t2 = x.^2;
t3 = y.^2;
t4 = z.^2;
name1 = t2+t3+t4;
if nargout > 1
    name2 = t2-t3-t4;
end

You can speed up the evaluation of a symbolic function at given coordinates by converting the symbolic function to an anonymous MATLAB function. Use matlabFunction to perform the conversion. Evaluation of a symbolic function returns symbolic numbers that are exact, while evaluation of a MATLAB function returns double-precision numbers.

Create a symbolic function f(x,y,z) that is a function of x, y, and z.

syms f(x,y,z)
f(x,y,z) = y*z*sin(x) + x*sin(z)*cos(y) - z^3;

Create 3-D grid coordinates at the specified intervals.

[xDouble,yDouble,zDouble] = meshgrid(1:20,1:50,1:20);

Evaluate the symbolic function at these coordinates. Measure the elapsed time using a pair of tic and toc calls.

tic
fResult = f(xDouble,yDouble,zDouble);
toc
Elapsed time is 1.582651 seconds.

Here, evaluation is slow, but it returns symbolic numbers that are exact. Show a sample of the results.

fResult(1:2,1:2,20)
ans = 

(20sin(1)+cos(1)sin(20)-800020sin(2)+2cos(1)sin(20)-800040sin(1)+cos(2)sin(20)-800040sin(2)+2cos(2)sin(20)-8000)

To speed up the evaluation of the function, convert the symbolic function to a MATLAB function using matlabFunction. Evaluate the MATLAB function at the same coordinates.

f1 = matlabFunction(f);
tic
fResult = f1(xDouble,yDouble,zDouble);
toc
Elapsed time is 0.010327 seconds.

Here, evaluation is faster. The evaluated MATLAB function returns double-precision numbers. Show a sample of the results.

fResult(1:2,1:2,20)
ans = 2×2
103 ×

   -7.9827   -7.9808
   -7.9667   -7.9644

Input Arguments

collapse all

Symbolic input to be converted to a MATLAB function, specified as a symbolic expression, function, vector, or matrix. When converting sparse symbolic vectors or matrices, specify the Sparse name-value argument as true.

Symbolic input to be converted to a MATLAB function with N outputs, specified as symbolic expressions, functions, vectors, or matrices, separated by commas.

matlabFunction does not create a separate output argument for each element of a symbolic vector or matrix. For example, ht = matlabFunction([x + 1, y + 1]) creates a MATLAB function with one output argument, while ht = matlabFunction(x + 1, y + 1) creates a MATLAB function with two output arguments.

Name-Value Arguments

Specify optional pairs of arguments as Name1=Value1,...,NameN=ValueN, where Name is the argument name and Value is the corresponding value. Name-value arguments must appear after other arguments, but the order of the pairs does not matter.

Example: matlabFunction(f,File='myfile',Optimize=false)

Before R2021a, use commas to separate each name and value, and enclose Name in quotes.

Example: matlabFunction(f,'File','myfile','Optimize',false)

Comments to include in the file header, specified as a character vector, cell array of character vectors, or string array.

Path to the file containing the generated MATLAB function, specified as a character vector or string scalar. The generated function accepts arguments of type double and can be used without Symbolic Math Toolbox. If File is empty, matlabFunction generates an anonymous function. If File does not end in .m, the function appends .m.

When writing to a file, matlabFunction optimizes the code using intermediate variables named t0, t1, and so on. To disable code optimization, use the Optimize argument.

See Write MATLAB Function to File with Comments.

Flag allowing optimization of code written to a function file, specified as true or false.

When writing to a file, matlabFunction optimizes the code by default using intermediate variables named t0, t1, and so on.

Using matlabFunction without the File argument (or with a file path specified as an empty character vector) creates a function handle. In this case, the code is not optimized by default. If you try to enforce code optimization by setting Optimize to true, then matlabFunction throws an error.

See Disable Code Optimization.

Flag that switches between sparse and dense matrix generation, specified as true or false. When you specify Sparse as true, the generated MATLAB function represents symbolic matrices with sparse numeric matrices. Specify Sparse as true when you convert symbolic matrices containing many zero elements. Often, operations on sparse matrices are more efficient than the same operations on dense matrices.

See Generate Sparse Matrices.

Order of input variables or vectors in a generated MATLAB function, specified as a character vector, a vector of symbolic variables, or a one-dimensional cell array of character vectors, symbolic variables, or vectors of symbolic variables.

The number of specified input variables must equal or exceed the number of symbolic variables in f. Do not use the same names for the input variables specified by Vars and the output variables specified by Outputs.

By default, when you convert symbolic expressions that contain only lowercase letters for the variable names, the order of the input variables is alphabetical. When you convert symbolic functions, their input arguments appear in front of other variables, and all other variables are ordered alphabetically.

Specify Vars as a vector to generate a MATLAB function with input arguments that are scalar variables. Specify Vars as a cell array to generate a MATLAB function with input arguments that are a combination of scalar and vector variables.

See Specify Input Arguments for Generated Function.

Names of output variables, specified as a one-dimensional cell array of character vectors.

If you do not specify the output variable names, then they coincide with the names you use when calling matlabFunction. If you call matlabFunction using an expression instead of individual variables, the default names of output variables consist of the word out followed by a number, for example, out3.

Do not use the same names for the input variables specified by Vars and the output variables specified by Outputs.

matlabFunction without the File argument (or with a file path specified as an empty character vector) creates a function handle. In this case, matlabFunction ignores the Outputs flag.

See Specify Output Variables.

Output Arguments

collapse all

Function handle that can serve as an input argument to numerical functions, returned as a MATLAB function handle.

Limitations

  • Some symbolic functions that have no corresponding MATLAB functions operating on the double data type, such as simplify and solve, are kept as symbolic functions in the converted MATLAB function handle or file. The converted file that consists of these functions cannot be deployed using MATLAB Coder™ or MATLAB Compiler™. You need to create your own functions with the double data type to replace these symbolic functions. If you are interested in a symbolic function that cannot be deployed, please contact MathWorks Technical Support.

Tips

Version History

Introduced in R2008b