Gudhi

-----------

Release date: September 2018

As a major new feature, the GUDHI library now offers a Python interface
to the Nerve and Graph Induced Complex. The GUDHI conda package is now available through conda-forge channel.

Below is a list of changes made since Gudhi 2.2.0:
- [Nerve and Graph Induced Complex](https://gudhi.inria.fr/python/latest/nerve_gic_complex_user.html) (new Python interface)
- Cover complexes, that provably contain topological information about the input data.

- Compilation issue (bug fix)
- CMake warning with ninja generator.
- thread_local management on old XCode versions.
- Boost dependency for Windows Python module.

- [GUDHI conda package](https://gudhi.inria.fr/conda/)
- The GUDHI conda package is now available through conda-forge channel.

-----------

Release date: June 2018

As a major new feature, the GUDHI library now offers a Čech complex module, a sparse version of the Rips complex and
a utility to build the Rips complex from a correlation matrix (no Python interface yet).

Below is a list of changes made since Gudhi 2.1.0:

- [Čech complex](https://gudhi.inria.fr/doc/latest/group__cech__complex.html) (new package)
- The Čech complex is a simplicial complex where the set of all simplices is filtered by the radius of their minimal enclosing ball.

- [Rips complex](https://gudhi.inria.fr/doc/latest/group__rips__complex.html) (new functions and interfaces)
- A sparse version of the Rips complex
- Rips complex from a correlation matrix utility

- [Dockerfile](https://gudhi.inria.fr/dockerfile/) (new installation process)
A Dockerfile example is now provided to compile, test and install GUDHI in a container

- CGAL 4.12 compilation issue (bug fix)

- CMake minimal version is now 3.1
- To take advantage of the latest features and simplify CMakeLists.txt files

-----------

Release date: January 2018

As a major new feature, the GUDHI library now offers persistence representations and cover complex (no Python interface yet).

Below is a list of changes made since Gudhi 2.0.1:

- [Cover complex](https://gudhi.inria.fr/doc/latest/group__cover__complex.html) (new package)
- Nerves and Graph Induced Complexes are cover complexes, that provably contain topological information about the input data.

- [Representations of persistence diagrams](https://gudhi.inria.fr/doc/latest/group___persistence__representations.html) (new package)
- It contains implementation of various representations of persistence
diagrams. It implements basic functionalities which are neccessary to use
persistence in statistics and machine learning.

- [Simplex tree](https://gudhi.inria.fr/doc/latest/group__simplex__tree.html) (new functions and interfaces)
- Graph expansion with a blocker oracle.
- Cech complex implementation example using CGAL mini spheres in fixed dimension.
- Automatic dimension set mechanism.

- Alpha complex (new function)
- [Weighted periodic 3D](https://gudhi.inria.fr/doc/latest/_alpha_complex_2weighted_periodic_alpha_complex_3d_persistence_8cpp-example.html) version utility.

- CGAL 4.11 compilation issue (bug fix)

- [Cubical complex](https://gudhi.inria.fr/doc/latest/group__cubical__complex.html) (bug fix)
- Perseus file read function.
- Computations of incidence indices between cubes.
- Missing periodic argument for the Python version.

- Documentation
- New [file formats](https://gudhi.inria.fr/doc/latest/fileformats.html) section.
- Bugs fix

- [Utilities](https://gudhi.inria.fr/utils)
- Separate examples from utilities.

- GUDHI Debian package is available for Debian Testing distribution.

-----------

Release date: September 2017

This minor GUDHI library version is fixing issues and improves persistence graphical tools module. All the new modules comes with their Python interface and a lot of examples, even in Python.

Below is a list of changes made since Gudhi 2.0.0:

- Spatial searching (new function)

- Spatial searching is now offering a radius search method.

- Persistence graphical tools (interfaces improvement)

- Add a band boot display mechanism on persistence diagram.
- Number of points and barcode limitation before the display.
- Interface modification to read persistence from a file.

- Bottleneck distance (bug fix)

- Read persistence files with infinity values bug is fixed.

- Persistent cohomology (bug fix)

- Weighted Alpha complex 3d persistence bug is fixed.

- Simplex tree (dead code)

- Remove useless global filtration attribute, getter and setter.

- cython

- Plot persistence functions improvement.
- Rename cythonize_gudhi.py in setup.py to be conform with Python conventions.
- Windows python module compilation issue fix.
- Documentation generation bug fix.
- Reduce CMake interactions.
- Rips complex memory leak fix.

- Doxygen

- Using MathJax.js to generate LaTeX instead of png.

- CMake

- Boost dependencies improvement.
- Conda compilation issues fix.
- Modules activation/desactivation mechanism for compilation and test.

- Data points generator

- Move data/points/generator into src/common/utilities.

-----------

Release date: April 2017

As a major new feature, the GUDHI library now offers an interface with Python. All the new modules comes with their Python interface and a lot of examples, even in Python.

Below is a list of changes made since Gudhi 1.3.1:

- Bottleneck distance (new package)

- Bottleneck distance measures the similarity between two persistence diagrams.

- cython (new package)

- A Cython package allows to compile a Python interface with the GUDHI library.

- Spatial searching (new package)

- Spatial searching is a wrapper around [CGAL dD spatial searching](http://doc.cgal.org/latest/Spatial_searching/index.html)
algorithms that provides a simplified API to perform (approximate) neighbor searches.

- Subsampling (new package)

- Subsampling offers methods to subsample a set of points.

- Tangential complex (new package)

- A Tangential Delaunay complex is a non filtered simplicial complex designed to
reconstruct a k-dimensional manifold embedded in d-dimensional Euclidean space.

- Witness complex (new relaxed version with a new interface)

- Witness complex Wit(W,L) is a simplicial complex defined on two sets of
points in Rd. The new relaxed version is filtrated. The new interface eases the
simplicial complex construction.

- Alpha complex (new interface)

- Alpha complex is a simplicial complex data structure constructed from the
finite cells of a Delaunay Triangulation. The new interface eases the simplicial
complex construction.

- Rips complex (new interface)

- The Rips complex is a simplicial complex constructed from an expanded one-skeleton
graph. The new interface eases the simplicial complex construction and allows the Rips
complex to be build from a distance matrix.

-----------

Release date: September 2016

Below is a list of changes made since Gudhi 1.3.0:

- Bug fix
As Simplex_handle default type was an 'int' in the simplex tree data structure, persistence cohomology computation was segmentation faulting
when the number of simplices was going further to the 'int' maximum value (2.7 billions of simplices on a classical modern machine and OS).
Simplex_handle default type is now an 'std::uint32_t' and can go up to about 4 billions of simplices.

- CMake
- The messages from CMake has been rewritten to be more consistent.

- Coding conventions
- CMake projects names and C++ namespaces have been homogenized.

- Documentation
- Mandatory and optional third party libraries have been separated in the documentation.

- Data sets
- in data/points/generator : thanks to [Aurélien Alvarez](http://www.aurelienalvarez.org/), aurelien_alvarez_surfaces_in_R8.py is
a script to generate points on a surface in R8.