This file documents software changes.

1.0.0 Satisfying CRAN checks.

0.9.0 Writing help and finalizing demos.

0.8.0 Other examples dealing with -omics (gene or protein) networks reverse-engineering were added.

0.7.5 Simulation functions.

0.7.0 Multicore and cluster examples for use with the new functions.

0.6.5 Updating existing demos for use with the new functions.

0.6.0 Several major improvements: code efficiency was enhanced further and the main function was split into several ones so that:

- it can be applied separately to different groups of variables to ensure high scalability or to analyze only some variables of the dataset (especially useful when dealing with omics datasets when focusing on some gene subsets or function)

- the simulation process can now be performed in several steps, if the user needs additionnal simulations either for some variables or for the whole dataset

- the random data generator or even the simulated datasets can be reused with different estimation process

0.5.0 Major code rewriting to improve performances. Code was profiled to improve performance even further. As a consequence, the package used for von Mises distribution fitting and random value generation was switched from movMF to Rfast.

0.4.0 Major functionality enhancement to allow for the use of generalized linear models such as glmnet models. Several selection functions are provided.

0.3.5 One demo was added for the adaptative lasso.

0.3.0 Major functionality enhancement to allow for custom grouping or selection functions, enabling the use, for instance, of adaptative lasso or generalized elastic net.

0.2.5 Two options were added to the boost function to return:

* either the list of simulated dataframes -to avoid memory problems with big datasets- or

* the result of the repeated variable selection process as a matrix with 0/1 values to allow for a customized analysis of those results.

0.2.0 Two demos were added for the regular lasso.

0.1.5 Some multicore functionality was enabled.

0.1.0 Package is functional with two preset grouping functions and several selection functions for the regular lasso or the regular elasticnet.