## Documentation Center |

On this page… |
---|

Specifying MATLAB Startup Options |

You can specify startup options (also called command flags or
command-line switches) that instruct the MATLAB^{®} program to perform
certain operations when you start it. On all platforms, you specify
the options as arguments to the `matlab` command
when you start at the operating system prompt. For example, the following
starts MATLAB and suppresses the display of the splash screen.

matlab -nosplash

On Windows^{®} platforms, you can precede a startup option
with either a hyphen (`-`) or a slash (`/`).
For example, `-nosplash` and `/nosplash` are
equivalent.

You can add selected startup options (also called command flags or switches for the command line) to the target path for your shortcut on the Windows platform for MATLAB. For more information about the options, see Commonly Used Startup Options.

To use startup options for the MATLAB shortcut icon on a Windows platform, follow these steps:

Right-click the shortcut icon for MATLAB and select

**Properties**from the context menu. The Properties dialog box for MATLAB opens to the**Shortcut**pane.In the

**Target**field, after the target path for`matlab.exe`, add the startup option, and click**OK**. For example, adding`-r "filename"`runs the MATLAB code file`filename`after startup.

This example instructs MATLAB to automatically run the
file `results` after startup, where `results.m` is
in the startup folder or on the search path for MATLAB. The statement
in the **Target** field might appear as:

C:\Program Files\MATLAB\R2010b\bin\matlab.exe -r "results"

Include the statement in double quotation marks (`"statement"`).
Use only the file name, not the file extension or path. For example, MATLAB produces
an error when you run

... matlab.exe -r "D:\results.m"

Use semicolons or commas to separate multiple statements. This
example changes the format to `short`, and then runs
the MATLAB code file `results`:

... matlab.exe -r "format('short');results"

Separate multiple options with spaces. This example starts MATLAB without
displaying the splash screen, and then runs the MATLAB code file `results`:

... matlab.exe -nosplash -r "results"

At startup, MATLAB automatically executes the file `matlabrc.m` and,
if it exists, `startup.m`. The file `matlabrc.m`,
which is in the `matlabroot``/toolbox/local` folder,
is reserved for use by MathWorks^{®} and by the system
administrator on multiuser systems.

The file `startup.m` is for you to specify
startup options. For example, you can modify the default search path,
predefine variables in your workspace, or define defaults for Handle Graphics^{®} objects.
Use the following statements in a `startup.m` file
to add the specified folder, `/home/username/mytools`,
to the search path, and to change the current folder to

addpath /home//mytools cd /home/username/mytoolsusername

Place the `startup.m` file in the default or
current startup folder, which is where MATLAB first looks for
it. For more information, see MATLAB Startup Folder.

The following table provides a list of some commonly used startup
options for both Windows and UNIX^{®} platforms. For more information,
including a complete list of startup options, see the `matlab (Windows)` or the `matlab (UNIX)` documentation.

You can pass Perl variables to MATLAB on startup by using the `-r` option
of the `matlab` function. For example, assume a MATLAB function `test` that
takes one input variable:

function test(x)

To pass a Perl variable instead of a constant as the input parameter,
follow these steps. This command starts MATLAB and runs `test` with
the input argument `10`.

Create a Perl script such as

#!/usr/local/bin/perl $val = 10; system('matlab -r "test(' . ${val} . ')"');

Invoke the Perl script at the prompt using a Perl interpreter.

For more information, see the `-r` option in the `matlab
(Windows)` or `matlab (UNIX)` documentation.

When MATLAB starts, it constructs the class path for Java software
using `librarypath.txt` as well as `classpath.txt`.
If you call Java software from MATLAB, see more about this
in The Java Class Path and Locating Native Method Libraries.

Was this topic helpful?