ParaMonte MATLAB release notes

This project follows Semantic Versioning. To access the latest release of the package, visit the ParaMonte GitHub repository release page or the ParaMonte page on MathWorks FileExchange central package repository.

Version 2.x.x

Version 2.5.0 – January 1, 2021

Major enhancements

• This release is a a major step toward further portability of the kernel routines of the ParaMonte::MATLAB library. The kernel library dependencies are now properly handled and recognized at runtime without such aggressive actions as permanently redefining the environmental path variables, most importantly, PATH and LD_LIBRARY_PATH on Linux/macOS.

• The ParaMonte::MATLAB library is now capable of recognizing the existing MPI libraries such as MPICH and OpenMPI on user’s system and avoid further installation of a new MPI library if it is deemed unnecessary.

• The ParaMonte kernel routines are now capable of handling user-input file paths that contain white-space (blank) or other exotic characters.

Minor enhancements

• As of this version, when the ParaMonte::MATLAB library is called in MATLAB -batch mode (from the command line) for the first time, the library avoids asking the user’s response to the question of installing an MPI library if it is missing on the user’s system. This will prevent undesired crashes of the simulations for the first time when the simulation is run from outside the MATLAB session. However, the onus will be on the user to ensure an MPI library exists on the system if they intend to run simulations in parallel.

• The ParaMonte::MATLAB library packages for different Operating systems and processor architecture are now separate from each other. This change was made to lower the overall size of ParaMonte::MATLAB by only keeping the relevant files in each packaging of the library. The current release contains three separate packages for ParaMonte::MATLAB,
  • libparamonte_matlab_windows_x64,
  • libparamonte_matlab_darwin_x64,
  • libparamonte_matlab_linux_x64.
• Typo-fixes in the documentation of the library.

MATLAB versions used for this release

• Windows: MATLAB 9.6.0.1072779 (R2019a)
• Linux: MATLAB 9.8.0.1323502 (R2020a)
• macOS: MATLAB 9.8.0.1323502 (R2020a)

MATLAB version compatibility

This release has been tested with MATLAB 2018, 2019, and 2020. It should be also compatible with MATLAB 2017, but is not tested. If you notice an incompatibility with any of the above MATLAB versions, please report this issue to the developers for a resolution at:

https://github.com/cdslaborg/paramonte/issues

Version 2.4.0 – December 23, 2020

• This version of the library was internal to the developers and not released to the public.

Version 2.3.0 – December 17, 2020

Major enhancements

• This update presents several major performance, accuracy, and verification enhancements to the ParaMonte kernel routines, in particular, to the ParaDRAM sampler.

• An extensive set of over 866 tests have been added that test all aspects of the ParaMonte kernel library.

• The issue of Windows file locking, that led to the occasional crashes of the ParaDRAM and ParaDISE simulations in multiChain parallelism mode, is now resolved.

• The ParaDRAM class in paramonte is now also available as Paradram and paradram, although the original label will remain the default preferred method of ParaDRAM object instantiation.

Version 2.2.1 – November 15, 2020

Minor enhancements

• Minor enhancements to the Kernel library build scripts and dependencies management.

• More informative error messages are now printed on MATLAB console if any error happens during the ParaMonte library setup on macOS for the first time.

Version 2.2.0 – October 29, 2020

Enhancements

• The cmake software dependency installation failure now does not nullify the installation of other dependencies.

• The IO debugging info of all ParaMonte samplers have been enhanced. In cases of wrong syntax or syntax-breaking input values in the simulation output files, the error messages are now more informative and point directly to the exact location of of error in the input file.

• The Integrated Autocorrelation (IAC) for sample refinement in ParaDRAM sampler of ParaMonte is now set to the average of all variables’ IAC values instead of the maximum IAC value. This will lead to less aggressive decorrelation of the final sample, which means significantly larger final sample sizes, without compromising the i.i.d. property of the final refined sample. This behavior can be reversed back to the original by specifying “max” or “maximum” along with the requested refinement method, SampleRefinementMethod = "batchmeans max" or SampleRefinementMethod = "BatchMeans-max" (case-insensitive).

Version 2.1.3 – October 15, 2020

Minor enhancements

• Further minor enhancements to the behavior of the checkForUpdate() method of the paramonte class.

Version 2.1.2 – October 15, 2020

Minor enhancements

• The checkForUpdate() method of the paramonte class now functions as expected.

Version 2.1.1 – October 9, 2020

Minor enhancements

• A Linux bug in the installation of the MPI library is now fixed.

Version 2.1.0 – October 3, 2020

Minor enhancements

• A new simulation specification overwriteRequested has been added to all ParaMonte samplers. If True and the ParaMonte sampler detects an existing set of old simulation output files in the output path of the current simulation with the same names as the output file names of the current simulation, then, the ParaMonte sampler will overwrite the existing simulation files.

Version 2.0.1 – September 26, 2020

Minor enhancements

• The guidelines for the installation of the MPI library on macOS have been improved.

• The minor bug in GridPlot class method rotateAxesLabels() that caused the readSample() , readChain(), readMarkovChain() to crash upon adding Grid plots is now fixed.

• The minor bug in the naming of the ParaMonte kernel library files on macOS (Darwin) is now fixed.

Version 2.0.0 – September 22, 2020

Major enhancements to the ParaMonte / ParaDRAM sampler interfaces

• The entire ParaMonte MATLAB interface library has been revamped. The new naming conventions, visualization, and computing tools are significantly nicer to deal with and in some cases, orders of magnitude faster than the previous major release.

• The simulation output files reading is now completely overhauled. In particular, the output file reader methods are now capable of handling input file paths that point to a directory. In such cases, it will search the input directory for files matching the requested file name pattern. If no input file is provided to the file reader methods, the current working directory will be search for the the potential simulation files that match the requested pattern.

• Several new post-processing functionalities have now been added, such as the ability to seamlessly parse the contents of the output *_report.txt, *_restart.txt, and *_progress.txt simulation files, in addition to the other output files (*_sample.txt and *_chain.txt) that could be parsed in the previous versions.

• The newly-added readRestart() method is now added to the ParaDRAM sampler class. User can now parse the contents of the output ASCII-format restart files. This is particularly useful to visualize the dynamics of the ParaDRAM sampler class, such as the evolution of the proposal distribution’s location, shape, and covariance matrix.

• The GridPlot() class now has two additional methods setAxesLabels() and setAxesLimits() which can directly set the labels and limits of axes, hassle-free.

Minor enhancements

• The single value assignment to spec.targetAcceptanceRate component of a ParaDRAM object is now properly handles. For example, the following code is valid as expected,

  import paramonte as pm
  pmpd = pm.ParaDRAM()
  pmpd.spec.targetAcceptanceRate = 0.23 # this is now valid
  pmpd.spec.targetAcceptanceRate = [0.2, 0.3] # this is also valid, which limits the acceptance rate to the specified range
  

• The default background color in all plots is now "white".
• The rotateAxisLabels() of the GridPlot() class is now renamed to rotateAxesLabels().

Bug fixes

• ParaDRAM readMarkovChain() no-output-option bug is now fixed. When calling readMarkovChain(), user can now either provide the output variable or not.

Version 1.x.x

Version 1.1.0 – June 5, 2020

• Enhancements and bug fixes to the kernel routines.
• Several major enhancements and bug fixes to the MATLAB kernel and interface routines.
• MatDRAM now supports fully-deterministic restart functionality.

Version 1.0.0 – June 1, 2020 – Initial release

• This is the first public release of the ParaMonte MATLAB library.

New features

• ParaDRAM sampler: Parallel Delayed-Rejection Adaptive Metropolis-Hastings Markov Chain Monte Carlo Sampler.
• ParaMonte Interface to the MATLAB Programming language.
• ParaMonte simulation-output visualization via the ParaMonte MATLAB interface.