Show all
Author Commit Message Labels Comments Date
Chris Park
removed automake *.am files merged Maria's changes added to skip *o files in .hgignore
Chris Park
add to ignore *.o files
Chris Park
modified a bug that used 'dmissing' flag checking with "_arg" instead of "_given"
Actually added PCL2Bin files
-Added memory mapping support for pcl files -Added the workaround for really long pcl lines -Added PCL2Bin tool
Chris Park
merge annotation OBO updates, plus PR/RECALL columns to DChecker
Chris Park
merge Hypergeometric clarifications
Chris Park
merge support of 'bin' binary PCL files
Chris Park
merged qdab support and additional functionalities for Dat2Dab
Chris Park
added .hgignore to ignore deps cache Po etc generated files
KEGG no longer wraps lines so a substantially larger buffer size is now needed for lines from the ko file.
Make memory mapping of output files optional via -M flag. This flag affects the inference mode and should only be used if you have sufficient local storage for intermediate networks. Without this flag, it is best to have sufficient RAM for intermediate networks.
Chris Park
Added new tool MIed, similar to MIer however calculates MI between two dataset only using the union of the two gene sets. In contrast, MIer uses the sum of all genes among all datasets given to calculate MI. MIed allows MI to be constant even after adding new dataset into the integration.
Curtis Huttenhower
Fix bug in CDat::Open( CDat, vector<CGenes>*, ... ) for importing known negatives Thanks to Maria Chikina! Fix bug in CMeasureImpl::MeasureTrim for reordering unpaired vectors Fix bug in CPCL::Open error message generation Add ability to OpenGene to deal correctly with changing line widths Fix default Answerer behavior with positive/negative gene sets Add Bonferroni-corrected p-value output to Clinician Improve formatting of gene set lookup output in Dab2Dad
Improve COALESCE termination criteria (slightly more permissive) Fix bug in missing value handling for aligned measures in CMeasure Fix bug for ties in rank transform in CPCL::RankTransform Add PValueSpearman, FisherTransform, better TCDF to CStatistics Fix bug in newline handling in COALESCE Add Spearman correlation option to Clinician Fix bug in default edge cutoff value in Cliquer (none by default) Add randomization scoring to DChecker Add gene set limitatio…
Divided the code execution in two different stages: training and prediction
Seperated the execution in two stages: Training and Learning
- Modified joint probability computation between the species. - Added a few user friendly command line arguments
Updated SpeciesConnector removing Gibbs sampling.
Fixed SpeciesConnector to execute mkdir in UNIX.
Fixed Dat2Dab.cpp: Test for -1 rather than 0 for checking missing values. Added new tool SpeciesConnector for performing cross-species inference.
Fix build scripts to pick up extlib dependencies Update cmdline files based on new build
Update documentation to include new email and Mercurial repo location Update Combiner to make random effects + AD normality the default Fix a scary-stupid memory leak (my fault) in CHalfMatrix::SetSize Update COntologyKEGG to parse more recent versions of the KEGG Ontology Add PCC normalization to CDat and Normalizer - thanks to Arjun Krishnan! Add minimum edge count between groups for inclusion to Funcifier Add CStatistics::AndersonDarlingScore and Test as a normal…
Add covariate mutual information approximation to Counter Approximates joint P(X1, X2) as sum(M, prod(X, P(X|M))P(M)) over covariate In our case, covariate is gene membership in GO/KEGG/etc. terms Works great! Regularizes Bayes nets at least as well as full MI Fix symmetric CDat normalization to use normal CDF
Curtis Huttenhower
Remove obsolete VS2005 project files Update Windows build for new Boost v1.42
rmcg:00408: merged changes to boost and build script
rmcg:00408: added hg update to nightly build/make script
Fix push hook script to remove temporary file, avoiding permissions errors
Add new boost regexp library to BOOST_LIBS along with graph library
adding script directory back in
  1. Prev
  2. Next