2011-08-31 kingaa * [r74] DESCRIPTION, NAMESPACE, R/deprecated.R, R/hansen.R, man/ouch-deprecated.Rd, tests/exacttree.R, tests/exacttree.Rout.save: - remove deprecated functions and options, at long last 2011-07-25 kingaa * [r73] inst/ChangeLog: - update changelog * [r72] DESCRIPTION, inst/ChangeLog, inst/NEWS, man/hansen.Rd: - fix typo in documentation of 'hansen': sqrt.alpha and sigma should have 'nchar*(nchar+1)/2' entries. Thanks to Roland Sookias for bringing this typo to my attention. - put NEWS file into standard format 2010-07-11 kingaa * [r69] DESCRIPTION, R/hansen.R, inst/ChangeLog: - an infelicity in the 'simulate' method for 'hansentree' objects was discovered by Philip Bergmann. There was an unwanted (inexplicable) dependence on the data in that the optima were recomputed from the data for each simulation rather than simply being taken from the 'theta' slot. This has been remedied: now the fitted optima are used in the simulations. Thanks to Philip for bringing this to light. 2010-06-23 kingaa * [r67] inst/ChangeLog: - update the changelog for version 2.6-2 2010-06-18 kingaa * [r66] R/aaa.R: - removed aaa.R: no longer needed * [r65] DESCRIPTION, R/brown.R, R/hansen.R, R/ouchtree.R, man/hansen.Rd: - minor cleanup of hansen example - use the Collate field in the DESCRIPTION file - get rid of 'R/aaa.R' * [r64] DESCRIPTION, man/hansen.Rd: - add multivariate example to 'hansen.Rd' 2010-01-21 kingaa * [r60] inst/ChangeLog: - version 2.6-1 2010-01-19 kingaa * [r59] man/hansen.Rd: - improve the documentation 2010-01-14 kingaa * [r58] R/hansen.R: - pass 'clist' to the optimizers to avoid breaking them * [r57] DESCRIPTION, R/aaa.R, R/hansen.R, R/methods.R, inst/NEWS, man/anolis.ssd.Rd, man/bimac.Rd, man/hansen.Rd, man/hansentree-methods.Rd, man/paint.Rd, tests/boot.R, tests/boot.Rout.save, tests/exacttree.R, tests/exacttree.Rout.save, tests/simtree.R, tests/simtree.Rout.save: - In the 'hansen' function, the use of the argument 'alpha' was confusing. This vector parameterizes the "selection" portion of the OU model, which since version 2 has been assumed to be a positive-definite matrix (corresponding to stabilizing selection). In particular, it contains the entries of the lower-triangular matrix square root of this positive definite matrix. To make the name more mnemonic, this argument has been renamed 'sqrt.alpha'. This is a superficial change meant to improve intelligibility: the algorithm itself has not changed. To prevent loss of backward compatibility, use of argument 'alpha' will continue to work, though it will generate a warning. In some future release, use of 'alpha' as an argument to 'hansen' will no longer generate a warning but will instead be passed through to the control list for 'optim' if method='optim' and will otherwise generate an error. 2009-10-07 kingaa * [r56] inst/ChangeLog: * [r55] man/hansen.Rd, man/hansentree-methods.Rd: - cosmetology 2009-08-11 kingaa * [r45] inst/ChangeLog: version 2.5-7 * [r44] DESCRIPTION, R/ouchtree.R: now, we use a non-recursive algorithm for constructing the lineages. not sure if this is superior: it does appear to be slightly faster on the examples. also, we now have some additional error checking for self-ancestral nodes. * [r43] man/hansen.Rd: minor fix 2009-07-19 kingaa * [r42] inst/ChangeLog: version 2.5-6 * [r41] man/hansentree-methods.Rd: fix codoc mismatch in argument list for plot-hansentree 2009-07-13 kingaa * [r40] DESCRIPTION: fixed CITATION typo * [r39] inst/CITATION: whoops. 2009-06-26 kingaa * [r38] DESCRIPTION: version 0.25-5 * [r37] man/hansen.Rd, man/hansentree-methods.Rd, man/ouch-package.Rd: minor improvements to documentation 2009-03-29 kingaa * [r36] R/hansen.R: pass the hessian option to subplex 2009-03-11 kingaa * [r35] DESCRIPTION: * [r34] inst/ChangeLog: * [r33] man/hansen.Rd, man/hansentree-methods.Rd: fix bug in documentation of alpha, sigma arguments give Details section explaining how the alpha and sigma matrices are parameterized * [r32] inst/CITATION: use pomp.r-forge.r-project.org as the URL for the project 2009-01-29 kingaa * [r31] inst/ChangeLog: * [r30] DESCRIPTION: * [r29] man/hansentree-methods.Rd: all arguments are now (nominally) documented. 2009-01-28 kingaa * [r28] man/hansentree-methods.Rd: 2009-01-27 kingaa * [r27] DESCRIPTION: * [r26] man/hansentree-methods.Rd: * [r25] man/hansentree-methods.Rd: 2009-01-25 kingaa * [r23] DESCRIPTION: minor documentation fixes * [r22] man/bimac.Rd, man/browntree-class.Rd, man/ouchtree-class.Rd: 2008-12-16 kingaa * [r21] DESCRIPTION, inst/ChangeLog: 2008-12-09 kingaa * [r20] inst/ChangeLog: * [r19] DESCRIPTION, R/hansen.R: When the Choleski decomposition fails (typically when the vcov matrix has a numerically zero eigenvalue), formerly an error was generated. Under these circumstances, the stationary distribution of the OU process does not exist and the likelihood is therefore zero. This fix traps the error and returns -Inf for the log-likelihood (actually +Inf for the deviance in the return from ou.lik.fn). Optimizers that are robust to Inf can therefore proceed. 2008-12-08 kingaa * [r18] inst/CITATION: use r-forge URL 2008-09-11 kingaa * [r17] inst/ChangeLog: * [r16] DESCRIPTION: fix glitch that occurs when 'data' is not in the same order as the nodes patch contributed by Jim Regetz and Jonathan Davies * [r15] R/brown.R, R/hansen.R, tests/exacttree.R, tests/exacttree.Rout.save: fix glitch that occurs when 'data' is not in the same order as the nodes patch contributed by Jim Regetz and Jonathan Davies * [r14] R/glssoln.R: cosmetic 2008-07-29 kingaa * [r13] inst/NEWS: 2008-07-28 kingaa * [r12] inst/ChangeLog: * [r11] DESCRIPTION: * [r10] tests/snizard.R, tests/snizard.Rout.save, tests/snizard_tree.ape, tests/squamates.csv: 2008-07-25 kingaa * [r9] R/paint.R, man/paint.Rd: fix bug in paint when 'tree' is a 'hansentree' * [r8] DESCRIPTION: * [r7] R/aaa.R, R/ape2ouch.R, R/brown.R, R/hansen.R, R/methods.R, R/ouchtree.R, R/paint.R, R/plot.R, R/rmvnorm.R: improve error messages 2008-07-23 kingaa * [r6] DESCRIPTION: 2008-07-22 kingaa * [r5] inst/ChangeLog: * [r4] inst/COPYING, inst/ChangeLog, inst/LICENSE, inst/OCHANGELOG: * [r3] R/plot.R, man/ouchtree-class.Rd: fix bug with plot options add 'legend' option to plot-ouchtree * [r2] DESCRIPTION, NAMESPACE, R/aaa.R, R/ape2ouch.R, R/brown.R, R/deprecated.R, R/glssoln.R, R/hansen.R, R/methods.R, R/ouchtree.R, R/paint.R, R/plot.R, R/rmvnorm.R, data, data/anolis.ssd.csv, data/bimac.csv, inst, inst/CITATION, inst/COPYING, inst/ChangeLog, inst/GPL, man/anolis.ssd.Rd, man/ape2ouch.Rd, man/bimac.Rd, man/brown.Rd, man/browntree-class.Rd, man/hansen.Rd, man/hansentree-class.Rd, man/hansentree-methods.Rd, man/ouch-deprecated.Rd, man/ouch-package.Rd, man/ouchtree-class.Rd, man/paint.Rd, src, src/covar-matrix.c, src/ouch.h, src/weight-matrix.c, tests, tests/boot.R, tests/boot.Rout.save, tests/exacttree.R, tests/exacttree.Rout.save, tests/simtree.R, tests/simtree.Rout.save, tests/snizard.R, tests/snizard.Rout.save, tests/snizard_tree.ape, tests/squamates.csv: initial transfer to rforge 2007-06-25 stefan7th * [r1] ., R, man: Email in Readme changed