Fixed bug in the bootstrap method for bumphunter. The coefficient for which inference was calculated was always 2 instead of the coefficient coef passed as an argument.
Fixed NAMESPACE issues.
nearestgene() now loads TT.rda into the local environment
known_transcripts(), and consequently matchGenes(), was yielding mostly NA for the Refseq / 'annotation' column (iff there were multiple refseq ids for the given Entrez/Gene id)
Fixed data/TT.rda, as the new value of known_transcripts().
Cleaned up NEWS.
Added GitHub URL to DESCRIPTION.
Cleaned up duplicates in Suggests:.
Added NEWS file.
Fixed a bug related to >= for numerics.
Added smoothing using a gaussian kernal as implemented in the locfit package through the function locfitByCluster.
Added closeSockets for cleanup for doParallel on Windows.
More bugfixes for windows; now using foreachCleanup().
Added a 'bumps' class and print method.
annotateNearest / regionMatch now give NA annotations for queries with no nearest subject (perhaps because the seqname is missing from the subject). Previously this was taken to be mistaken input and a hard error raised.
Speedup of fwer computations using foreach.
Added boundedClusterMaker.
bug fix to internal function .getModT (which are not used in the main bumphunter functions). Now the t-statistics returned a correct.