Performance and Memory

Profile code, improve performance, reduce memory requirements

Write your code to be simple and readable, especially for the first implementation. Code that is prematurely optimized can be unnecessarily complex without providing a significant gain in performance. Then, if speed is an issue, you can measure how long your code takes to run and profile your code to identify bottlenecks. If necessary, you can take steps to improve performance.

MATLAB^® handles data storage for you automatically. However, if memory is an issue, you can identify memory requirements and apply techniques to use memory more efficiently.

Apps

Profiler

Run code and measure execution time to improve performance

Functions

expand all

Measure and Profile Code

`tic`	Start stopwatch timer
`toc`	Read elapsed time from stopwatch
`cputime`	CPU time used by MATLAB
`timeit`	Measure time required to run function
`profile`	Profile execution time for functions
`bench`	MATLAB benchmark

Identify and Reduce Memory Requirements

`memory`	Memory information
`inmem`	Names of functions, MEX files, and classes in memory

Cache Results of Function Calls

`memoize`	Add memoization semantics to function handle
`MemoizedFunction`	Call memoized function and cache results
`clearAllMemoizedCaches`	Clear caches for all `MemoizedFunction` objects

Topics

Measure and Profile Code

Measure the Performance of Your Code
Use the timeit function or the stopwatch timer functions, tic and toc, to time how long your code takes to run.
Profile Your Code to Improve Performance
Use the Profiler to measure the time it takes to run your code and identify which lines of code consume the most time or which lines do not run.
Measure Code Complexity Using Cyclomatic Complexity
Quantify code complexity based on cyclomatic complexity.

Improve Performance

Techniques to Improve Performance

To speed up the performance of your code, there are several techniques that you can consider.
- Preallocation
- Vectorization

Identify and Reduce Memory Requirements

How MATLAB Allocates Memory
Write more memory-efficient code by understanding how MATLAB allocates memory.
Strategies for Efficient Use of Memory
Reduce memory usage in your programs, use appropriate data storage, avoid fragmenting memory, and reclaim used memory.
Avoid Unnecessary Copies of Data
MATLAB can apply memory optimizations when passing function inputs by value.
Resolve “Out of Memory” Errors
Troubleshoot errors when MATLAB cannot allocate the requested memory.

Related Information

Large Files and Big Data