------------------- Version: 0.7 Date: 2021-02-13 * Fix matrix object having more than one class in R 4.0.0. Clean up the rest of methods. ------------------- Version: 0.6-9 Date: 2020-07-13 * Remove Bioconductor packages from the suggested list to fix the issue on macOS ------------------- Version: 0.6-8 Date: 2020-06-22 * Fix matrix object having more than one class in R 4.0.0 ------------------- Version: 0.6-7 Date: 2020-05-07 * Fix CRAN issues: * Replace all require() with requireNamespace() * use namespace tags for kohonen:: and Mclust:: functions * Add importFrom() to NAMESPACE file ------------------- Version: 0.6-6 Date: 2014-03-24 * Declared 'biocLite' as a global variable in source code ------------------- Version: 0.6-5 Date: 2014-03-21 * Moved 'clValid' vignette to folder 'vignettes' * Moved 'methods' and 'class' from 'Depends' field to 'Imports' field in DESCRIPTION file ------------------- Version: 0.6-4 Date: 2011-10-17 * Added 'clValid.pdf' back to 'inst/doc' ------------------- Version: 0.6-3 Date: 2011-10-17 * Removed unnecessary files from 'inst/doc' * Changed license to LGPL-3 ------------------- Version: 0.6-2 Date: 2011-01-25 * Fixed bug in NAMESPACE file (removed call to .onLoad()) ------------------- Version: 0.6-1 Date: 2010-11-05 * Changed format for reading biological annotation files from Excel 2003/2007 to comma separated * Split R code into four files: clValid-Classes.R, clValid-Methods.R, clValid-internal.R, clValid-functions.R ------------------- Version: 0.6 Date: 2009-11-17 * matchGO : Made change to identify annotated probes with category="all" argument. * clValid : Missing values are now allowed in clValid(). For the correlation metric the use="pairwise.complete.obs" option is used, while the na.rm=TRUE option is used when calculating means for the AD and ADM measures. * annotationListToMatrix : Added new function to convert annotation lists to TRUE/FALSE logical matrices: columns = functional classes, rows = genes * readAnnotationFile : Added function to read biological annotation file from Excel spreadsheet * getRanksWeights : Added function to obtain ranks and weights for rank aggregation from clValid object * BSI, BHI : Changed functions to work with TRUE/FALSES matrices from annotateListToMatrix (thanks to Rainer Machne for suggestion) * sota : History of diversities for ancestor nodes are now stored. * sota : Moved break for maxDiversity with unrest.growth==FALSE to occur at same point as break for maxCycles (thanks to Rainer Machne) * clValid : Added argument 'verbose' to track pecentage completion of cluster validation * clValid : Now attempts automatic download and installation of Bioconductor annotation packages if not currently installed * clValid-class : Added slot 'labels' to contain the character vector giving the item (gene) labels. * clValid, BHI, BSI : Updated to use package GO.db instead of defunct GO package. Also, added argument 'dropEvidence' to select type of GO evidence used. * Packages RankAggreg and RODBC added to list of recommended packages * Packages kohonen and mclust changed from required to recommended list of packages * Packages GO and moe430 removed from list of recommend packages, added packages GO.db and moe430a.db * Vignette updated to include example of rank aggregation using package RankAggreg, and also illustrates usage of 'dropEvidence' argument and reading in biological annotation files using Excel ------------------- Version: 0.5-7 Date: 2008-07-28 * Added latex style file and made minor changes to clValid-class.Rd ------------------- Version: 0.5-6 Date: 2008-01-27 * sota, sota.print, sota.plot : Fixed these to print and plot properly when actual number of clusters is fewer than requested number (i.e. some clusters are not populated) * Added fixes to catch cases where fewer than requested number of clusters is returned by clustering method. A warning message is given when this occurs. ------------------- Version: 0.5-4 Date: 2007-12-14 * The default titles for plotting the validation measures were changed to "Internal validation", "Stability validation", and "Biological validation" for internal, stability, and biological validation, respectively. * The functions for calculating each validation measure: dunn, connectivity, BHI, BSI, stability; and the internal function matchGO are all exported now. Users can call these directly if they have an external clustering algorithm they want to evaluate, or if they just have a dissimilarity matrix (though the original data matrix is still required for the stability measures APN, AD, ADM, and BSI). * clValid : Now accepts data frames w/integer data * clValid : Attempts to automatically install Bioconductor metadata packages if they are not currently available * clValid : A new argument 'maxitems' was added to limit the number of clustered items. If R is being used interactively and 'maxitems' is exceeded, the user is prompted about whether to continue. If R is not being used interactively the function exits with an error message. * clValid : Accepts ExpressionSet objects rather than the now defunct exprSet objects