[root]/src/dr/math/distributions
Author | Changes | Lines of Code | Lines per Change |
---|---|---|---|
Totals | 152 (100.0%) | 7823 (100.0%) | 51.4 |
alexei.drummond | 23 (15.1%) | 3232 (41.3%) | 140.5 |
msuchard | 53 (34.9%) | 1849 (23.6%) | 34.8 |
Sibon.Li@gmail.com | 10 (6.6%) | 756 (9.7%) | 75.6 |
rambaut | 23 (15.1%) | 552 (7.1%) | 24.0 |
baele.guy@gmail.com | 4 (2.6%) | 406 (5.2%) | 101.5 |
alexander.alekseyenko | 3 (2.0%) | 335 (4.3%) | 111.6 |
jheled | 12 (7.9%) | 257 (3.3%) | 21.4 |
denise.kuehnert@gmail.com | 1 (0.7%) | 231 (3.0%) | 231.0 |
jen.tom | 1 (0.7%) | 118 (1.5%) | 118.0 |
dong.w.xie@gmail.com | 10 (6.6%) | 41 (0.5%) | 4.1 |
mdhall272@gmail.com | 2 (1.3%) | 21 (0.3%) | 10.5 |
dong.w.xie | 8 (5.3%) | 21 (0.3%) | 2.6 |
tbcbedford@gmail.com | 1 (0.7%) | 3 (0.0%) | 3.0 |
akaruiws | 1 (0.7%) | 1 (0.0%) | 1.0 |
Fixed log determine of a matrix and positive-definite check in WishartDistribution; there is still an error in LUPDecomposition.isPD()
5 lines of code changed in 2 files:
Implemented a scaled mixture of Wishart distribution for precision matrix inference.
72 lines of code changed in 6 files:
Added a half-T-distribution for more flexible specification of precision matrices in trait evolution.
53 lines of code changed in 1 file:
Implemented a log-determinant routine for numerical stability and fixed getParameterValue in DiagonalMatrix
31 lines of code changed in 1 file:
Fixed log determinate for large diagonal matrices; refactored transforms.
50 lines of code changed in 2 files:
Keep Colt flag off unless shown to provide more reliable simulants.
7 lines of code changed in 2 files:
- nextGamma(alpha, lambda) in Colt needs to be called with 1/scale instead of scale
- also replaced log(x/scale) by log(x) - log(scale) to avoid numerical errors when evaluating very small values from a gamma distribution
23 lines of code changed in 1 file:
Side-stepping trouble with Gamma RV generation
16 lines of code changed in 1 file:
Updated Compound DirichletMultinomial distribution to use both the rates and frequencies + dispersion parametrization. New example XML included.
90 lines of code changed in 2 files:
Allowed to take infinite maximum or minimum values
21 lines of code changed in 2 files:
Added log transformed kernel density estimation to the list
1 lines of code changed in 2 files:
Kernel density estimation for densities with bounded support by using log transformation
382 lines of code changed in 1 file:
Porting 'evaluateLate' interface method from 1.7 branch.
4 lines of code changed in 1 file:
trunk: merge prior dialog and distribution changes to trunk
29 lines of code changed in 9 files:
Fix to last commit.
1 lines of code changed in 1 file:
TruncatedDistribution now adjusts its bounds to be inside the source distribution's (i.e., if it is an OffsetPositive). LogNormalDistribution had a lower bound of -Inf, now set to zero.
9 lines of code changed in 2 files:
Runtime errors galore.
19 lines of code changed in 2 files:
Initial, horribly slow, inefficient re-write of antigenic likelihood.
5 lines of code changed in 1 file:
trunk: merge
12 lines of code changed in 1 file:
Fix for Issue 581 - NormalDistribution was hanging when asked for a CDF when stdev was 0. Now uses better CDF function (that was already implemented) that has bound checking.
4 lines of code changed in 1 file:
Fixed bug in calculation of log density in BiasedMultivariateDiffusionModel. Branches should properly scale with time now. This seems to be working.
3 lines of code changed in 1 file:
Oops. Messed up implementing logPDF for UniformDistribution.
1 lines of code changed in 2 files:
Trying to fix up the confusion with uniform bounds. The parameter object in beauti now has 'uniformUpper' and 'uniformLower' rather than using the truncationUpper/Lower (uniform prior cannot be truncated now). Changed the logPDF function of UniformDistribution to take the log before the reciprocal to reduce chance of underflow (note however that taking the log of Double.MAX_VALUE (or twice that if the lower bound is -MAX_VALUE) may still fail). Testing now being undertaken...
6 lines of code changed in 1 file:
Can specify non-integer degrees-of-freedom for Wishart distributions
11 lines of code changed in 1 file:
Fixed integer overflow (when prior sampling) and mixed-math in RandomWalkOperator
1 lines of code changed in 1 file:
Fixed prior graphs in BEAUti, inverse gamma cdf and started stochastic Dollo
4 lines of code changed in 1 file:
Seem to have got the Antigenic MDS working correctly with threshold assay values and replicate measurements.
34 lines of code changed in 1 file:
An approximation to the tail of a normal distribution added. To be tested...
73 lines of code changed in 1 file:
Removed commons-maths-2.0.jar from lib folder
160 lines of code changed in 1 file:
TruncatedDistribution more robustly handles quantiles.
41 lines of code changed in 3 files:
Generalized TruncatedDistribution that wraps another distribution. Quantile functions are not robust.
11 lines of code changed in 1 file:
Some major refactoring of PriorDialog. Probably horribly broken but at least it compiles.
126 lines of code changed in 1 file:
Cleaned up PriorDialog to better render distributions if possible.
50 lines of code changed in 3 files:
Bounds checking on Dirchlet distribution
6 lines of code changed in 2 files:
Tidy up more msatBMA related things.
1 lines of code changed in 1 file:
231 lines of code changed in 1 file:
Trunk: big change of new Tracer, and this check-in starts using List instead array in Variate.
20 lines of code changed in 7 files:
Started MV Polya Likelihood model
245 lines of code changed in 1 file:
removed unmappable character
35 lines of code changed in 1 file:
Bifractional diffusion is a pain
3 lines of code changed in 2 files:
(51 more)