| NEWS | R Documentation |
The packages Biobase and CHGbase are now imported (instead of suggested) in the DESCRIPTION-file to avoid errors on CRAN.
Modification of input checks to meet CRAN's best practice.
Citations are updated.
The packages Biobase and CHGbase are now suggested in the DESCRIPTION-file to avoid errors on CRAN.
The .Rd-file of the package is updated.
Citations are updated.
A typo corrected.
Some warnings on C++-code tackled.
The functions motifStatsVAR1 and pruneMotifStats added.
Dependency on the marray-package removed (often give an error in CRAN checks). Consequently, the color scheme employed in the evaluateVAR1fit is less refined.
More extensive description added to the description field of the description file.
Correct package name inserted in registerDynamicSymbol.c.
R code aligned for nicer viewing on Github.
The hpvP53mir data object is replaced by the correct version.
Several typo's in manuals corrected.
More consistent wording across manuals introduced.
Citation updated.
Functions momentS, ridgePchordal, optPchordal, and support4ridgep are now imported from the rags2ridges-package.
Three new modules are added: fused ridge estimation of multiple VAR(1) models, and ridge estimation of the VAR(2) and VARX(1) models.
Bug repaired in the graphVAR1-function. Input combination type="globalPC" and prune=TRUE produced a warning.
Some code from Rcpp-functions replaced by slightly more efficient code.
Meaningless code from Rcpp-functions removed.
Citation updated.
More data from the same experiment have been added. Documentation modified accordingly. Of note: the original data have been renamed to hpvP53rna.
A misnomer in the output and documentation of the nodeStatsVAR1-function fixed.
The formula of the VAR(1) model is added to documentation of the ridgeVAR1-function.
Some re-organization of the src-directory.
Rcpp-functions made exportable to other packages.
Input checks of the ridgeVAR1-function contained inconsistencies: repaired.
Function momentS added: evaluates the moments of the sample (inverse) covariance matrix.
Change to the description and namespace-file. Importing now from methods, for compatibility with Bioconductor 3.4.
A bug repaired in the ridgeVAR1-function. The option-combination fitA="ss" and a non-empty matrix being passed to the argument zerosA gave an error (an argument was not supplied to an internal C++ function).
A bug repaired in the ridgeVAR1-function. When specifying the option-combination fitA="ss" and non-empty matrices being passed to the arguments zerosA and zerosP the wrong order of arguments was supplied to an internal C++ function.
Extra argument (TburnIn) for dataVAR1-function. It specifies number of burn-in time points before data is sampled from the process.
The impulseResponseVAR1-function and its arguments changed slightly. Previously, the function aggregated all impulse responses between a variate at the current time point and those at the T future time points. The current implemention evaluates the impulse response in each variate at the current time point t on those at time point t+T. Apologies for inconvenience caused.
The mutualInfoVAR1-function and its arguments changed slightly. Previously, the function aggregated all mutual informations between a variate at the current time point and those at the T future time points. The current implemention evaluates the mutual information between each variate at the current time point t and those at time point t+T. Apologies for inconvenience caused.
The nodeStatsVar1-function changed internally to accomodate the changes in the impulseResponseVAR1 and mutualInfoVAR1-functions.
Extra argument (zeros) for the sparsifyVAR1-function. It specifies which elements in the auto-regression coefficient matrix are known to zero and should not be taken along in the sparsification.
Added a NEWS file.
Modified the example code of the evaluateVAR1fit. In this version, the evaluateVAR1fit function is no longer executed in the example code. This avoids saving .eps files during the package checking.