December 2007 Commit Log

Number of Commits:
11
Number of Active Developers:
3
msuchard 2007-12-19 07:49 Rev.: 642

Fixed (what I think was) a critical bug in the trace analysis. If user does not specify a burnin, then the trace log reader sets the burnin to 10%. But, after that, TraceAnalysis would reset it to -1 (ooops). I was seeing some odd results where the mean fell way outside of the 95% HPD -- the initial values were very large and should have been discarded.

114 lines of code changed in 1 file:

msuchard 2007-12-12 07:31 Rev.: 641

AncestralStateTreeLikelihood now samples "?" ambiguity code on tree tips.

176 lines of code changed in 1 file:

alexei.drummond 2007-12-11 18:10 Rev.: 640

Added Graham Jones' density for root time in birth-death model (when sample proportion is 1.0). Lets test it :-)

287 lines of code changed in 1 file:

alexei.drummond 2007-12-11 17:29 Rev.: 639

Added Graham Jones' comments on Yule model prior

567 lines of code changed in 4 files:

Michael.DefoinPLatel 2007-12-10 14:19 Rev.: 637

Fix the RatechangeLikelihood to get the autocorrelated clock model working. Rates can be absolute or relative.

217 lines of code changed in 4 files:

Michael.DefoinPLatel 2007-12-10 14:16 Rev.: 636

Add an option in LogAnalyser to export the standard error

22 lines of code changed in 2 files:

Michael.DefoinPLatel 2007-12-10 14:14 Rev.: 635

Display the log likelihood and all negative numbers in decimal format

2 lines of code changed in 1 file:

msuchard 2007-12-10 12:46 Rev.: 634

Added generalized GTR model with separate root and stationary distributions for irreversible processes. Added a Dirchlet prior distribution for informative root or stationary distributions. NodeTraitLogger just annotates trees with all traits.

527 lines of code changed in 9 files:

msuchard 2007-12-05 08:02 Rev.: 631

MultivariateTraitLikelihood now takes a specially normalized BranchRateModel. ContinuousEpochBranchRateModel returns such a model.

171 lines of code changed in 4 files:

msuchard 2007-12-02 16:16 Rev.: 630

Tried to fix an oversight in SubtreeSlideOperator in which only univariate traits with the name "trait" get swapped when the root changes. All traits, regardless of name or dimension should be swapped. Please check that TreeModel.swapAllTraits() handles this correctly. This will be important for the phenotype models (like the multivariate diffusion processes). Also started to get the Brownian diffusion model ready for inhomogeneity in time by rescaling. Being single really SUCKS ass, but I sure do have a lot more time to code.

612 lines of code changed in 5 files:

msuchard 2007-12-01 05:47 Rev.: 629

Added restricted growth function output for clusterings in InhibitionAssayLikelihood. Ultimately this class should become a sub-class of something like ClusterLikelihood.

430 lines of code changed in 2 files:

November 2007 »

Generated by StatSVN 0.7.0