https://github.com/numericalalgorithmsgroup/dfbgn
Python solver for large-scale nonlinear least-squares minimization without derivatives
https://github.com/numericalalgorithmsgroup/dfbgn
least-squares nonlinear-optimization numerical-analysis numerical-methods numerical-optimization optimization optimization-algorithms python scientific-computing
Last synced: about 2 months ago
JSON representation
Python solver for large-scale nonlinear least-squares minimization without derivatives
- Host: GitHub
- URL: https://github.com/numericalalgorithmsgroup/dfbgn
- Owner: numericalalgorithmsgroup
- License: gpl-3.0
- Created: 2020-07-17T08:37:06.000Z (over 5 years ago)
- Default Branch: master
- Last Pushed: 2021-11-03T01:06:42.000Z (over 4 years ago)
- Last Synced: 2026-01-10T18:15:37.832Z (about 2 months ago)
- Topics: least-squares, nonlinear-optimization, numerical-analysis, numerical-methods, numerical-optimization, optimization, optimization-algorithms, python, scientific-computing
- Language: Python
- Homepage: https://numericalalgorithmsgroup.github.io/dfbgn/
- Size: 6.43 MB
- Stars: 10
- Watchers: 1
- Forks: 2
- Open Issues: 2
-
Metadata Files:
- Readme: README.rst
- License: LICENSE.txt
Awesome Lists containing this project
README
=========================================
DFBGN: Derivative-Free Block Gauss-Newton
=========================================
.. image:: https://github.com/numericalalgorithmsgroup/dfbgn/actions/workflows/python_testing.yml/badge.svg
:target: https://github.com/numericalalgorithmsgroup/dfbgn/actions
:alt: Build Status
.. image:: https://img.shields.io/badge/License-GPL%20v3-blue.svg
:target: https://www.gnu.org/licenses/gpl-3.0
:alt: GNU GPL v3 License
.. image:: https://img.shields.io/pypi/v/dfbgn.svg
:target: https://pypi.python.org/pypi/dfbgn
:alt: Latest PyPI version
DFBGN is a Python package for nonlinear least-squares minimization, where derivatives are not available.
It is particularly useful when evaluations of the objective are expensive and/or noisy, and the number of variables to be optimized is large.
This is an implementation of the algorithm from our paper: C. Cartis, L. Roberts, Scalable Subspace Methods for Derivative-Free Nonlinear Least-Squares Optimization, (2021) [`preprint `_].
DFBGN is based on `DFO-LS `_, but is better-suited when there are many variables to be optimized (so the linear algebra in DFO-LS is too slow).
Unlike DFO-LS, DFBGN does not currently support bound constraints on the variables.
If you are interested in solving general optimization problems (without a least-squares structure), you may wish to try `Py-BOBYQA `_.
Citation
--------
If you use DFBGN in a paper, please cite:
Cartis, C., and Roberts, L., `Scalable Subspace Methods for Derivative-Free Nonlinear Least-Squares Optimization `_, *arXiv preprint arXiv:2102.12016*, (2021).
Requirements
------------
DFBGN 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/)
Installation using pip
----------------------
For easy installation, use `pip `_ as root:
.. code-block:: bash
$ [sudo] pip install dfbgn
or alternatively *easy_install*:
.. code-block:: bash
$ [sudo] easy_install dfbgn
If you do not have root privileges or you want to install DFBGN for your private use, you can use:
.. code-block:: bash
$ pip install --user dfbgn
which will install DFBGN in your home directory.
Note that if an older install of DFBGN is present on your system you can use:
.. code-block:: bash
$ [sudo] pip install --upgrade dfbgn
to upgrade DFBGN to the latest version.
Manual installation
-------------------
Alternatively, you can download the source code from `Github `_ and unpack as follows:
.. code-block:: bash
$ git clone https://github.com/numericalalgorithmsgroup/dfbgn
$ cd dfbgn
DFBGN is written in pure Python and requires no compilation. It can be installed using:
.. code-block:: bash
$ [sudo] pip install .
If you do not have root privileges or you want to install DFBGN for your private use, you can use:
.. code-block:: bash
$ pip install --user .
instead.
To upgrade DFBGN to the latest version, navigate to the top-level directory (i.e. the one containing :code:`setup.py`) and rerun the installation using :code:`pip`, as above:
.. code-block:: bash
$ git pull
$ [sudo] pip install . # with admin privileges
Testing
-------
If you installed DFBGN manually, you can test your installation by running:
.. code-block:: bash
$ python setup.py test
Alternatively, the HTML documentation provides some simple examples of how to run DFBGN.
Examples
--------
Examples of how to run DFBGN are given in the `documentation `_, and the `examples `_ directory in Github.
Uninstallation
--------------
If DFBGN was installed using *pip* you can uninstall as follows:
.. code-block:: bash
$ [sudo] pip uninstall dfbgn
If DFBGN 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.