Desbordante

Hello everyone! This release primarily focuses on bug fixes and improvements related to Numerical Association Rule (NAR) discovery. Here are the key changes:

* **Optional Seed Parameter:** Introduced an optional seed parameter in the NAR discovery algorithm to ensure repeatability across subsequent invocations.
* **Consistent NAR Discovery Example:** Updated the NAR discovery example to guarantee consistent results across different machines and invocation.
* **Improved Python Bindings:** Fixed the IND, AIND, and SFD Python bindings to prevent further invocations from disrupting previously found pattern instances.
* **CMake Version Requirement:** The README now clearly states the requirement for CMake version 3.25.
* **Build System Upgrade:** Transitioned from Make to Ninja for faster build times.
* **MacOS Performance Improvements:** Several algorithms (FDep, FastADC, FastOD) have received a significant speed boost, achieving performance parity with their Linux counterparts.

This release only fixes pypi.org metadata.

The changelog of v.2.3.0 can be seen [here](https://github.com/Desbordante/desbordante-core/releases/tag/v2.3.0).

Hello everyone! We are excited to share that with this release, we've added support for macOS, making it possible to install the Desbordante-core pip package on macOS via PyPi for CPython versions 3.8 through 3.13, as well as for PyPy versions 3.7 through 3.10. Additionally, we have introduced support for CPython 3.13 on Linux. This release also includes several important updates and new features:

* **Numerical Association Rule Discovery:** We have added support for numerical association rule discovery, a pattern specifically designed for extracting knowledge from data. Currently, we include an approximate discovery algorithm based on an evolutionary approach, which may miss some rules. A usage example is provided for reference.

* **EulerFD Algorithm:** A novel algorithm, EulerFD, has been introduced for the approximate discovery of exact functional dependencies. This algorithm is very fast but may miss some dependencies and introduce additional (non-existing) ones. To the best of our knowledge, it is currently the fastest existing approximate algorithm for this task. Two usage examples for approximate functional dependency discovery algorithms are included.

* **Approximate Denial Constraints:** We have added support for the discovery of approximate denial constraints. This recently developed pattern allows us to extract facts from data expressed by a Boolean formula. Discovery is performed using the FastADC algorithm, which is currently the fastest algorithm for this task. A usage example is provided.

* **Parallelization of HyFD Algorithm:** The HyFD algorithm, one of the most efficient algorithms for exact functional dependency discovery, has been parallelized. As a result, we have achieved up to a 10x improvement in running times.

* **Improved Differential Dependency Discovery Algorithm:** The SPLIT differential dependency discovery algorithm has been enhanced for improved performance, resulting in faster execution times (up to 17x). Also, its memory consumption has been significantly reduced (25x), allowing it to process larger datasets.

This release only fixes pypi.org metadata.

The changelog of v.2.2.0 can be seen [here](https://github.com/Desbordante/desbordante-core/releases/tag/v2.2.0).

Almost a year passed from our initial release of the core package. Over this time we have experienced a great deal of interest, getting 18K downloads and 400 stars. Therefore, we have decided to make this release a bit special by trying to introduce as many new interesting patterns into Desbordante as possible. While there are several of them, the obvious star of this release is matching dependencies—a pattern which will greatly help in data deduplication, data cleaning and many other data quality tasks.

Overall, this release contains pattern implementations accumulated in over half a year. We hope you will find it useful!

Changes:
* Added discovery of matching dependencies—a very expressive type of dependency, capable of capturing subtle inconsistencies in data by using various matching functions.
* Added discovery of many types of approximate functional dependencies. Before, we defined the error value of an approximate FD to be calculated using the g1 metric. Our new definition permits use of any error metric, as the alternative metrics are currently gaining popularity. Therefore, we are expanding the number of supported metrics in Desbordante and in this release we added discovery for the $\mu+$, $\tau$, $pdep$, and $\rho$ metrics.
* Added discovery of soft functional dependencies and corellations.
* Added validation of variable heterogeneous denial constraints.
* Added discovery and validation of approximate inclusion dependencies (using the $g^{‘}_3$ metric). Inclusion dependencies can help users to recover foreign keys, or to find joinable columns in a table or a collection of tables. Supporting an approximate version of this pattern will allow users to perform these tasks even when dealing with dirty data.
* Added validation of probabilistic functional dependencies.
* Examples were reorganized into three categories: 1) basic, which showcase a single pattern, 2) advanced, which illustrate various pattern nuances, and 3) expert, which demonstrate instances of complex programs that aim to provide tangible benefits for end-users by solving real-life problems using pattern discovery or validation.

Miscellaneous:
* Added HPIValid algorithm for discovery of UCCs. To the best of our knowledge, it is currently the most performant algorithm for this task, therefore we made it the default one.
* Added examples for UCC and AUCC mining.
* Fixed the AR example and added the support output to the Python bindings.
* Fixed lifetime issues with FD and UCC objects.

All novel patterns are coming with usage examples. Please note that the console version of Desbordante will be updated a bit later.

Release Notes

This minor release serves as a necessary step for isolating code of the console interface and moving it into a separate [repository](https://github.com/Desbordante/desbordante-cli). Our final goal is to create a dedicated Python package called [desbordante-cli](https://www.pepy.tech/projects/desbordante-cli), which will be implemented purely in Python. It will depend on the core [desbordante](https://www.pepy.tech/projects/desbordante) package that contains the C++ code for pattern mining and validation.

As such, we plan to make minor releases of the core package in the future, followed by the console ones. These releases will contain fewer features, but will come out a lot more frequently. The idea here is to make a release as soon as each individual algorithm is ready rather than accumulating several of them as we did previously. Once a sufficient number of features have been accumulated, a major release will be published, primarily for promotion purposes. It will not provide any new functionality, but will include all the accumulated changes since the last major release.

Changes:
* We have added support for a novel class of algorithms — the dynamic ones. The idea is to track changes in the dataset in order to update their result on-the-fly rather than processing the whole table again. As a result, they can be up to several orders of magnitude faster than classic (static) ones in some situations. Along with devising dynamic infrastructure, we have implemented the first dynamic algorithm — a dynamic functional dependency validator. A Python interface and an example are provided.
* We have added support for discovery of differential dependencies. Differential dependency is a relatively novel type of pattern which is very handy for detecting a particular relationship between two column sets. It can be seen as an extension of functional dependency which works well on dirty data. See the article about the pattern for more information. Its implemented discovery algorithm (Split) comes with a Python interface and an example.
* Discovery of association rules is now available via the Python and console interfaces. An example is also available.

Miscellaneous:
* Greatly improved the metric functional dependency verification example.
* Added approximate inclusion dependency discovery algorithms to the C++ core. Python interface, console interface, and an example are still in development.
* Fixed Python bindings for association rules: the AR objects can be properly copied now.
* Extended simple statistics module with ten string-related statistics; they are available via the Python interface.
* Fixed a CLI-breaking bug related to the CFD discovery algorithm.
* Improved column type deduction in the C++ core.