DFO-LS: Derivative-Free Optimizer for Least-Squares
DFO-LS is a flexible package for solving nonlinear least-squares minimization, without requiring derivatives of the objective. It is particularly useful when evaluations of the objective function are expensive and/or noisy. DFO-LS is more flexible version of DFO-GN.
This is an implementation of the algorithm from our paper: C. Cartis, J. Fiala, B. Marteau and L. Roberts, Improving the Flexibility and Robustness of Model-Based Derivative-Free Optimization Solvers, ACM Transactions on Mathematical Software, 45:3 (2019), pp. 32:1-32:41 [preprint]. For reproducibility of all figures in this paper, please feel free to contact the authors.
If you are interested in solving general optimization problems (without a least-squares structure), you may wish to try Py-BOBYQA, which has many of the same features as DFO-LS.
Documentation
See manual.pdf or here.
Citation
If you use DFO-LS in a paper, please cite:
Cartis, C., Fiala, J., Marteau, B. and Roberts, L., Improving the Flexibility and Robustness of Model-Based Derivative-Free Optimization Solvers, ACM Transactions on Mathematical Software, 45:3 (2019), pp. 32:1-32:41.
Requirements
DFO-LS requires the following software to be installed:
- Python 2.7 or Python 3 (http://www.python.org/)
Additionally, the following python packages should be installed (these will be installed automatically if using pip, see Installation using pip):
- NumPy 1.11 or higher (http://www.numpy.org/)
- SciPy 0.18 or higher (http://www.scipy.org/)
- Pandas 0.17 or higher (http://pandas.pydata.org/)
Optional package: DFO-LS versions 1.2 and higher also support the trustregion package for fast trust-region subproblem solutions. To install this, make sure you have a Fortran compiler (e.g. gfortran) and NumPy installed, then run pip install trustregion. You do not have to have trustregion installed for DFO-LS to work, and it is not installed by default.
Installation using pip
For easy installation, use pip as root:
$ [sudo] pip install DFO-LS
or alternatively easy_install:
$ [sudo] easy_install DFO-LS
If you do not have root privileges or you want to install DFO-LS for your private use, you can use:
$ pip install --user DFO-LS
which will install DFO-LS in your home directory.
Note that if an older install of DFO-LS is present on your system you can use:
$ [sudo] pip install --upgrade DFO-LS
to upgrade DFO-LS to the latest version.
Manual installation
Alternatively, you can download the source code from Github and unpack as follows:
$ git clone https://github.com/numericalalgorithmsgroup/dfols $ cd dfols
DFO-LS is written in pure Python and requires no compilation. It can be installed using:
$ [sudo] pip install .
If you do not have root privileges or you want to install DFO-LS for your private use, you can use:
$ pip install --user .
instead.
To upgrade DFO-LS to the latest version, navigate to the top-level directory (i.e. the one containing setup.py) and rerun the installation using pip, as above:
$ git pull $ [sudo] pip install . # with admin privileges
Testing
If you installed DFO-LS manually, you can test your installation by running:
$ python setup.py test
Alternatively, the HTML documentation provides some simple examples of how to run DFO-LS.
Examples
Examples of how to run DFO-LS are given in the documentation, and the examples directory in Github.
Uninstallation
If DFO-LS was installed using pip you can uninstall as follows:
$ [sudo] pip uninstall DFO-LS
If DFO-LS was installed manually you have to remove the installed files by hand (located in your python site-packages directory).
Bugs
Please report any bugs using GitHub's issue tracker.
License
This algorithm is released under the GNU GPL license. Please contact NAG for alternative licensing.
