Module related to Bayesian approach to inverse problems
Project description
The Bayesian inversion (edwin) package provides algorithm developed during my scientific research work in numerical computation for inverse problems (signal, image processing). Feel free to use them as you want. Any comments and contributions are welcome.
The name edwin is in reference to Edwin T. Jaynes, a great Bayesian Analysis scientific.
Acknowledgements
The use of edwin software package should be explicitly acknowledged in publications in the following form:
an acknowledgment statement: “Some of the results in this paper have been derived using some of the edwin package algorithms From F. Orieux et al. published in citations.
at the first reference, a footnote placed in the main body of the paper referring to the edwin web site, currently http://bitbucket.org/forieux/edwin
The citations are mentioned in documentation, References section of this file and are available in bibtex file.
Info
Author: François Orieux
Contact: orieux at iap dot fr
Project homepage: http://bitbucket.org/forieux/edwin
Downloads page: https://bitbucket.org/forieux/edwin/downloads
Contents
- improcessing.py
A module that implement the algorithm described in [2] for unsupervised myopic image deconvolution. However the myopic part is not actually available.
- inversion.py
A module that implement the algorithm described in [1] and use in [3-4] and other papers. It’s implement an unsupervised general inverse problem algorithm estimation, based on MCMC algorithm.
- sampling.py
Implementation of stochastic sampling algorithm, specially [1].
- optim.py
A module that implement classical optimisation algorithm for use of other module. They are design for very large system resolution (dim > 1e6).
Requirements
This package depends on my free otb package (utility functions).
Numpy version >= 1.4.1
otb version >= 0.2.1
Installation
The pip version:
pip install edwin
If you have not pip, download the archive, decompress it and to install in your user path, run in a command line:
python setup.py install --user
or for the system path, run as root:
python setup.py install
Development
This package follow the Semantic Versionning convention http://semver.org/. To get the development version you can clone the mercurial repository available here http://bitbucket.org/forieux/edwin
The ongoing development depends on my research activity but is open. I try to fix bugs.
License
edwin is free software distributed under the MIT license, see LICENSE.txt
References
A bibtex file is provided in the archive.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file edwin-0.2.0.tar.gz
.
File metadata
- Download URL: edwin-0.2.0.tar.gz
- Upload date:
- Size: 4.0 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 96318855268e9c01c8930d07f57e116968e9c50ea2033982ff6bd5c1ac08fb01 |
|
MD5 | 5adf9ff5f13662d112cbb0c1591c4814 |
|
BLAKE2b-256 | caaf4222d679ef0338997529d8441ab7700b4f0f2cea81ba6a2772682510bc22 |
File details
Details for the file edwin-0.2.0-py2-none-any.whl
.
File metadata
- Download URL: edwin-0.2.0-py2-none-any.whl
- Upload date:
- Size: 19.9 kB
- Tags: Python 2
- Uploaded using Trusted Publishing? No
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | dbbbcff8ce5e11a56305774fb78a55748f58084d6c5bd2dad4d0f3aeefd4317e |
|
MD5 | 032f8cd9402c1b119fac4337605465b5 |
|
BLAKE2b-256 | 6183c6ffcc2de71954c771381171aa5a5fa27f1a8bdcfd5586012664693b1e40 |