This topic shows how to share variables between workspaces or allow them to persist between function executions.
In most cases, variables created within a function are local variables known only within that function. Local variables are not available at the command line or to any other function. However, there are several ways to share data between functions or workspaces.
The most secure way to extend the scope of a function variable is to use function input and output arguments, which allow you to pass values of variables.
For example, create two functions, update1
and update2
,
that share and modify an input value. update2
can
be a local function in the file update1.m
, or can
be a function in its own file, update2.m
.
function y1 = update1(x1) y1 = 1 + update2(x1); function y2 = update2(x2) y2 = 2 * x2;
Call the update1
function from the command
line and assign to variable Y
in the base workspace:
X = [1,2,3]; Y = update1(X)
Y = 3 5 7
A nested function has access to the workspaces of all functions
in which it is nested. So, for example, a nested function can use
a variable (in this case, x
) that is defined in
its parent function:
function primaryFx x = 1; nestedFx function nestedFx x = x + 1; end end
When parent functions do not use a given variable, the variable
remains local to the nested function. For example, in this version
of primaryFx
, the two nested functions have their
own versions of x
that cannot interact with each
other.
function primaryFx nestedFx1 nestedFx2 function nestedFx1 x = 1; end function nestedFx2 x = 2; end end
For more information, see Nested Functions.
When you declare a variable within a function as persistent, the variable retains its value from one function call to the next. Other local variables retain their value only during the current execution of a function. Persistent variables are equivalent to static variables in other programming languages.
Declare variables using the persistent
keyword
before you use them. MATLAB® initializes persistent variables
to an empty matrix, []
.
For example, define a function in a file named findSum.m
that
initializes a sum to 0
, and then adds to the value
on each iteration.
function findSum(inputvalue) persistent SUM_X if isempty(SUM_X) SUM_X = 0; end SUM_X = SUM_X + inputvalue;
When you call the function, the value of SUM_X
persists
between subsequent executions.
These operations clear the persistent variables for a function:
clear all
clear
functionname
Editing the function file
To prevent clearing persistent variables, lock the function
file using mlock
.
Global variables are variables that you can access from functions or from the command line. They have their own workspace, which is separate from the base and function workspaces.
However, global variables carry notable risks. For example:
Any function can access and update a global variable. Other functions that use the variable might return unexpected results.
If you unintentionally give a “new” global variable the same name as an existing global variable, one function can overwrite the values expected by another. This error is difficult to diagnose.
Use global variables sparingly, if at all.
If you use global variables, declare them using the global
keyword
before you access them within any particular location (function or
command line). For example, create a function in a file called falling.m
:
function h = falling(t) global GRAVITY h = 1/2*GRAVITY*t.^2;
Then, enter these commands at the prompt:
global GRAVITY
GRAVITY = 32;
y = falling((0:.1:5)');
The two global statements make the value assigned to GRAVITY
at
the command prompt available inside the function. However, as a more
robust alternative, redefine the function to accept the value as an
input:
function h = falling(t,gravity)
h = 1/2*gravity*t.^2;
Then, enter these commands at the prompt:
GRAVITY = 32; y = falling((0:.1:5)',GRAVITY);
The evalin
and assignin
functions
allow you to evaluate commands or variable names from character vectors
and specify whether to use the current or base workspace.
Like global variables, these functions carry risks of overwriting existing data. Use them sparingly.
evalin
and assignin
are
sometimes useful for callback functions in graphical user interfaces
to evaluate against the base workspace. For example, create a list
box of variable names from the base workspace:
function listBox figure lb = uicontrol('Style','listbox','Position',[10 10 100 100],... 'Callback',@update_listBox); update_listBox(lb) function update_listBox(src,~) vars = evalin('base','who'); src.String = vars;
For other programming applications, consider argument passing and the techniques described in Alternatives to the eval Function.