The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.

FMM 0.4.1

Enhancements:

• Improved convergence behavior of the fitFMM_back() significantly reduces the number of iterations needed to reach a good solution.
• Clarified the warning message issued when high amplitude values are detected.

Changes:

• Fixed a bug in the construction of the logarithmic omega grid: exp(log(omegaMin)) did not match omegaMin exactly due to floating point precision. The grid is now constructed using L + 2 points, and the first and last are removed.
• Internal code restructuring: calculateSingleCosPhi() was separated from calculateCosPhi(), and a new function calculateDesignMatrix() was added.
• The show() method for FMM objects is now registered by default. The previous manual call to addShowMethod() is no longer needed.

New features: - Added a new internal function: mobius(), used in the updated optimization routine. - Added dependency on the gsignal package (used for the hilbert() function).

FMM 0.4.0

Enhancements:

• Improved computational efficiency: The grid search has been optimized to avoid redundant calculations, and a profile likelihood approach has been incorporated into the post-optimization procedure. These refinements lead to a more efficient convergence, reducing the need for explicit parallelization (parallelize argument) and iterative grid refinement (numReps argument).

• Refinements in plotFMM(): Bug fixes and improvements in prediction computation, now performed on a denser, evenly spaced grid in addition to the original time points.

• Code simplifications: The arguments numReps and parallelize have been deprecated, and the default value of showProgress has been changed from TRUE to FALSE.

New features:

• Additional arguments in fitFMM(): New arguments omegaMin, omegaMax, and omegaGrid have been added to provide more flexibility.

• New warnings messages: Additional warnings notify users about excessively large amplitude estimates and significant gaps in time points.

Changes:

• Repository update: The GitHub repository link in DESCRIPTION has been updated to FMMGroupVa/FMM.

• Vignette updates: The vignette has been revised to align with the latest modifications.

• Internal functions restructuring:

  • Functions bestStep1() and getApply() have been relocated from FMM_internal to FMM_internal_restr.
  • The previous version of step1FMM() has been renamed step1FMMRestr() and moved to FMM_internal_restr.
  • Two new internal functions, checkArguments() and checkSolution(), have been introduced to improve input validation and ensure the correctness of computed results.

FMM 0.3.1

Minor improvements and fixes:

• A vignette has been added.
• The default value of the argument showTime of fitFMM() function has been changed to FALSE.
• R2() and alwaysFalse() functions are listed in NAMESPACE file.
• A link to the GitHub repository has been added to DESCRIPTION file.
• An example for plotting the original time scale on the x-axis of the ggplots has been added to the documentation of the plotFMM() function.
• Add automated unit tests for fitFMM() function.
• fitFMM() prints informative error messages when the number of observation is insufficient or data are constant.

FMM 0.3.0

Enhancements:

• Optimization of the estimation procedure for the monocomponent, multicomponent and restricted FMM models. An embedded parallelized procedure is available for all the models for faster estimations.
• Code has been refined: it is now more object-oriented, code duplicity has been reduced, unnecessary orders have been suppressed, API usability has been improved, and the documentation is now automatically created with roxygen2.

New features:

• In addition to the estimation by blocks previously available for the restricted model, an exact implementation has been added for more accurate fits in expense of more computational time.

FMM 0.2.0

New features:

• Optimized monocomponent and multicomponent model computation.

FMM 0.1.2

Enhancements:

• Added a slot in the FMM object with information on the number of iteraction of the fitting process.
• Renamed the FMMPeaks function to getFMMPeaks.
• Renamed the generate_FMM function to generateFMM. Also parameters a, b and w to alpha, beta and omega, respectively.

New features:

• New show() method to print basic information on an FMM object.

FMM 0.1.1

Fixed the issues pointed out in the review of the CRAN submission:

• Modified the beginning of the description.
• Added details about internal functions in the documentation.
• Replaced the use of cat() by warning().
• options(warn=-1) is no longer used.
• No example in the documentation of the functions uses more than two processor cores.

FMM 0.1.0

• Submitted version 0.1.0 to CRAN.

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.