Skip to content

Latest commit

 

History

History
444 lines (380 loc) · 21 KB

README.md

File metadata and controls

444 lines (380 loc) · 21 KB

numbbo/coco: Comparing Continuous Optimizers

This code reimplements the original Comparing Continous Optimizer platform, now rewritten fully in ANSI C with other languages calling the C code. As the name suggests, the code provides a platform to benchmark and compare continuous optimizers, AKA non-linear solvers for numerical optimization. Languages currently available are

  • C/C++
  • Java
  • MATLAB/Octave
  • Python

Contributions to link further languages (including a better example in C++) are more than welcome.

For more information,

Requirements

  1. For a machine running experiments
  • A C compiler, such as gcc
  • Python >=2.6 with setuptools installed
  • optional: git
  1. For a machine running the post-processing
  • Python 2.6 or 2.7 with numpy (preferably >=1.7) and matplotlib installed. We recommend to install the Anaconda Python 2.7 library. Python 3 is not yet supported with the post-processing part of NumBBO/CoCO!

Windows Specifics

Under Windows, two alternative compile toolchains can be used:

  1. Cygwin which comes with gcc and make, available in 32- and 64-bit versions.
  2. MinGW's gcc (http://www.mingw.org/), which only comes in 32-bit, but also runs on 64-bit machines.

For using git under Windows (optional), we recommend installing TortoiseGit.

Language Specifics

Additional requirements for running an algorithm in a specific language.

  • C: make, such as GNU make (GNU make for Windows).
  • Java: gcc and any Java Development Kit (JDK), such that javac and javah are accessible (i.e. in the system path).
  • MATLAB: at least MATLAB 2008, for details, see here
  • Python on Windows with MinGW: Python 2.7 and the Microsoft compiler package for Python 2.7 containing VC9, available here. These are necessary to build the C extensions for the Python cocoex module for Windows. The package contains 32-bit and 64-bit compilers and the Windows SDK headers.
  • Octave: Octave 4.0.0 or later. On operating systems other than Windows, earlier versions might work.

Guaranties (None)

We tested the framework on Mac OSX, Ubuntu linux, Fedora linux, and Windows (XP, 7, 10) in various combinations of 32-bit and 64-bit compilers, python versions etc. Naturally, we cannot guarantee that the framework runs on any combination of operating system and software installed. In case you experience some incompatibilies, check out the Known Issues / Trouble Shooting Section below. Otherwise we will be happy if you can document them in detail on the issue tracker.

Getting Started

  1. Check out the Requirements above.

  2. Download the COCO framework code from github,

  • either by clicking here and unzip the zip file,
  • or (preferred) by typing git clone https://github.com/numbbo/coco.git. This way allows to remain up-to-date easily (but needs git to be installed). After cloning, git pull keeps the code up-to-date with the latest release.

CAVEAT: this code is still under heavy development. The record of official releases can be found here. The latest release corresponds to the master branch as linked above.

  1. In a system shell, cd into the coco or coco-<version> folder (framework root), where the file do.py can be found. Type, i.e. execute, one of the following commands once
  python do.py run-c
  python do.py run-java
  python do.py run-matlab
  python do.py run-octave
  python do.py run-python

depending on which language shall be used to run the experiments. run-* will build the respective code and run the example experiment once. The build result and the example experiment code can be found under code-experiments/build/<language> (<language>=matlab for Octave). python do.py lists all available commands.

  1. On the computer where experiment data shall be post-processed, run
  python do.py install-postprocessing

to (user-locally) install the post-processing. From here on, do.py has done its job and is only needed again for updating the builds to a new release.

  1. Copy the folder code-experiments/build/YOUR-FAVORITE-LANGUAGE and its content to another location. In Python it is sufficient to copy the file example_experiment.py. Run the example experiment (it already is compiled, in case). As the details vary, see the respective read-me's and/or example experiment files:

If the example experiment runs, connect your favorite algorithm to Coco: replace the call to the random search optimizer in the example experiment file by a call to your algorithm (see above). Update the output result_folder, the algorithm_name and algorithm_info of the observer options in the example experiment file.

Another entry point for your own experiments can be the code-experiments/examples folder.

  1. Now you can run your favorite algorithm on the bbob-biobj (for multi-objective algorithms) or on the bbob suite (for single-objective algorithms). Output is automatically generated in the specified data result_folder.

6. Postprocess the data from the results folder by typing

```
python -m bbob_pproc [-o OUTPUT_FOLDERNAME] YOURDATAFOLDER [MORE_DATAFOLDERS]
```

The name bbob_pproc will become cocopp in future. Any subfolder in the folder arguments will be searched for logged data. That is, experiments from different batches can be in different folders collected under a single "root" YOURDATAFOLDER folder. We can also compare more than one algorithm by specifying several data result folders generated by different algorithms.

A folder, ppdata by default, will be generated, which contains all output from the post-processing, including a ppdata.html file, useful as main entry point to explore the result with a browser. Data might be overwritten, it is therefore useful to change the output folder name with the -o OUTPUT_FOLDERNAME option.

For the single-objective bbob suite, a summary pdf can be produced via LaTeX. The corresponding templates in ACM format can be found in the code-postprocessing/latex-templates folder. LaTeX templates for the multi-objective bbob-biobj suite will follow in a later release. A basic html output is also available in the result folder of the postprocessing (file templateBBOBarticle.html).

  1. Once your algorithm runs well, increase the budget in your experiment script, if necessary implement randomized independent restarts, and follow the above steps successively until you are happy.

If you detect bugs or other issues, please let us know by opening an issue in our issue tracker at https://github.com/numbbo/coco/issues.

Description by Folder

  • the do.py file in the root folder is a tool to build the entire distribution. do.py is a neat and simplifying replacement for make. It has switches for just building some languages etc, type

      python do.py

    to see a list of all available commandes.

  • the code-experiments/build folder is to a large extend the output folder of the ./do.py build command.

    • the exampleexperiment.??? files in the build folder are the entry points to understand the usage of the code (as end-user). They are supposed to actually be executable (in case, after compilation, which should be taken care of by do.py and/or make) and run typically random search on (some of) the provided benchmark suites.

  • documentation and examples might not be too meaningful for the time being, even though code-experiments/documentation/onion.py describes a (heavily) used design pattern (namely: inheritance) in a comparatively understandable way (though the implementation in C naturally looks somewhat different). In the future, documentation will be contained mainly in the docs/ subfolder with the source code extracted automatically into pdfs in this folder and to web pages under the numbbo.github.io/ domain.

  • the code-experiments/src folder is where most of the important/interesting things happen. Many files provide comparatively decent documentation at the moment which are translated via doxygen into a more readable web page at numbbo.github.io/coco-doc/C/. Generally:

    • coco.h is the public interface, in particular as used in the example_experiment.c file
    • coco_internal.h provides the type definition of coco_problem_t
    • coco_suite.c is code that deals with an entire benchmark suite (i.e. a set of functions, eg. sweeping through them etc...)
    • coco_generics.c is somewhat generic code, e.g. defining a function call via coco_evaluate_function etc
    • coco_problem.c is the implementation of the coco_problem_t type/object (allocation etc).
    • observer / logger files implement data logging (as wrappers around a coco problem inheriting thereby all properties of a coco problem)
    • most other files implement more or less what they say, e.g. the actual benchmark functions, transformations, benchmark suites, etc.
    • currently, three benchmark suites and corresponding logging facilities are implemented:
      • bbob: standard single-objective BBOB benchmark suite with 24 noiseless, scalable test functions
      • bbob-biobj: a bi-objective benchmark suite, combining 10 selected functions from the bbob suite, resulting in 55 noiseless functions
      • toy: a simple, probably easier-to-understand example for reading and testing

  • code-experiments/tools are a few meta-tools, mainly the amalgamate.py to merge all the C code into one file

  • code-experiments/test contains unit- and integration-tests, mainly for internal use

  • code-postprocessing/bbob_pproc contains the postprocessing code, written in python, with which algorithm data sets can be read in and the performance of the algorithms can be displayed in terms of data profiles, ERT vs. dimension plots, or simple tables.

  • code-postprocessing/latex-templates contains LaTeX templates for displaying algorithm performances in publisher-conform PDFs for the GECCO and CEC conferences (for the single-objective bbob suite only, templates for the bi-objective bbob-biobj suite will be provided in a later release).

  • docs should contain an updated version of the documentation, see above.

  • howtos contains a few text files with internal howtos.

Known Issues / Trouble-Shooting

Java

javah call fails

If you see something like this when running python do.py run-java or build-java under Linux

COPY    code-experiments/src/coco.h -> code-experiments/build/java/coco.h
WRITE   code-experiments/build/java/REVISION
WRITE   code-experiments/build/java/VERSION
RUN     javac CocoJNI.java in code-experiments/build/java
RUN     javah CocoJNI in code-experiments/build/java
Traceback (most recent call last):
  File "do.py", line 590, in <module>
    main(sys.argv[1:])
  File "do.py", line 563, in main
    elif cmd == 'build-java': build_java()
  File "do.py", line 437, in build_java
    env = os.environ, universal_newlines = True)
  File "/..../code-experiments/tools/cocoutils.py", line 34, in check_output
    raise error
subprocess.CalledProcessError: Command '['locate', 'jni.h']' returned non-zero exit status 1

it means javah is either not installed (see above) or cannot be found in the system path, see this and possibly this for a solution.

Matlab

Path to matlab

If you see something like this when running python do.py build-matlab

AML	['code-experiments/src/coco_generics.c', 'code-experiments/src/coco_random.c', 'code-experiments/src/coco_suite.c', 'code-experiments/src/coco_suites.c', 'code-experiments/src/coco_observer.c', 'code-experiments/src/coco_runtime_c.c'] -> code-experiments/build/matlab/coco.c
COPY	code-experiments/src/coco.h -> code-experiments/build/matlab/coco.h
COPY	code-experiments/src/best_values_hyp.txt -> code-experiments/build/matlab/best_values_hyp.txt
WRITE	code-experiments/build/matlab/REVISION
WRITE	code-experiments/build/matlab/VERSION
RUN	matlab -nodisplay -nosplash -r setup, exit in code-experiments/build/matlab
Traceback (most recent call last):
  File "do.py", line 447, in <module>
    main(sys.argv[1:])
  File "do.py", line 429, in main
    elif cmd == 'build-matlab': build_matlab()
  File "do.py", line 278, in build_matlab
    run('code-experiments/build/matlab', ['matlab', '-nodisplay', '-nosplash', '-r', 'setup, exit'])
  File "/Users/auger/workviasvn/newcoco/numbbo/code-experiments/tools/cocoutils.py", line 68, in run
    universal_newlines=True)
  File "//anaconda/lib/python2.7/subprocess.py", line 566, in check_output
    process = Popen(stdout=PIPE, *popenargs, **kwargs)
  File "//anaconda/lib/python2.7/subprocess.py", line 710, in __init__
    errread, errwrite)
  File "//anaconda/lib/python2.7/subprocess.py", line 1335, in _execute_child
    raise child_exception
OSError: [Errno 2] No such file or directory

It might be because your system does not know the matlab command. To fix this, you should edit the file /etc/paths and add the path to the matlab bin file (Linux/Mac) or add the path to the folder where the matlab.exe lies to your Windows path. For instance, the etc/paths should look like something like this

/usr/local/bin
/usr/bin
/bin
/usr/sbin
/sbin
/Applications/MATLAB_R2012a.app/bin/

SMS-EMOA example does not compile under Mac

With the more complex SMS-EMOA example. The problem is related to the compilation of the external C++ hypervolume calculation in hv.cpp.

A fix for this issue consists in adding to the files hv.cpp and paretofront.c

#define char16_t UINT16_T

just before the line:

#include "mex.h"

Access to mex files denied

If it happens that you get some Access is denied errors during python do.py build-matlab or python do.py run-matlab like this one

C:\Users\dimo\Desktop\numbbo-brockho>python do.py run-matlab
Traceback (most recent call last):
  File "do.py", line 649, in <module>
    main(sys.argv[1:])
  File "do.py", line 630, in main
    elif cmd == 'run-matlab': run_matlab()
  File "do.py", line 312, in run_matlab
    os.remove( filename )
WindowsError: [Error 5] Access is denied: 'code-experiments/build/matlab\\cocoEv
aluateFunction.mexw32'

a reason can be that a previously opened Matlab window still has some file handles open. Simply close all Matlab windows (and all running Matlab processes if there is any) before to run the do.py command again.

Octave

none for the moment

Python

setuptools is not installed

If you see something like this

$ python do.py run-python  # or build-python
[...]
PYTHON  setup.py install --user in code-experiments/build/python
ERROR: return value=1
Traceback (most recent call last):
 File "setup.py", line 8, in <module>
   import setuptools
ImportError: No module named setuptools

Traceback (most recent call last):
 File "do.py", line 562, in <module>
   main(sys.argv[1:])
 File "do.py", line 539, in main
   elif cmd == 'build-python': build_python()
 File "do.py", line 203, in build_python
   python('code-experiments/build/python', ['setup.py', 'install', '--user'])
 File "/vol2/twagner/numbbo/code-experiments/tools/cocoutils.py", line 92, in p                                         ython
   universal_newlines=True)
 File "/usr/local/lib/python2.7/subprocess.py", line 575, in check_output
   raise CalledProcessError(retcode, cmd, output=output)
subprocess.CalledProcessError: Command '['/usr/local/bin/python', 'setup.py', 'i                                        nstall', '--user']' returned non-zero exit status 1

then setuptools needs to be installed:

    pip install setuptools

or easy_install setuptools should do the job.

Compilation During Install of cocoex Fails (under Linux)

If you see something like this:

$ python do.py run-python  # or build-python
[...]
cython/interface.c -o build/temp.linux-i686-2.6/cython/interface.o
cython/interface.c:4:20: error: Python.h: file not found
cython/interface.c:6:6: error: #error Python headers needed to compile C extensions, please install development version of Python.
error: command 'gcc' failed with exit status 1

Under Linux

  sudo apt-get install python-dev

should do the trick.

Module Update/Install Does Not Propagate

We have observed a case where the update of the cocoex Python module seemed to have no effect. In this case it has been successful to remove all previously installed versions, see here for a few more details.

Details

  • The C code features an object oriented implementation, where the coco_problem_t is the most central data structure / object. coco.h, example_experiment.c and coco_internal.h are probably the best pointers to start to investigate the code (but see also here). coco_problem_t defines a benchmark function instance (in a given dimension), and is called via coco_evaluate_function.

  • Building, running, and testing of the code is done by merging/amalgamation of all C-code into a single C file, coco.c, and coco.h. (by calling do.py, see above). Like this it becomes very simple to include/use the code in different projects.

  • Cython is used to compile the C to Python interface in build/python/interface.pyx. The Python module installation file setup.py uses the compiled interface.c, if interface.pyx has not changed. For this reason, Cython is not a requirement for the end-user.

  • We continuously test the code through the open source automation server Jenkins on one ubuntu 12.04 machine, one OSX 10.9 machine, and one 32-bit Windows 7 machine with cygwin.

Links and Documentation