1.7.8 ----- Attempt to check if PTHREAD_STACK_MIN is defined 1.7.7 ----- Include <limits.h> on pthreads builds 1.7.5 ----- Fix bug when scale estimate provided to rcModelPLM() 1.7.3/1.7.4 ----------- Fix bug with floating point argument provided where integer expected. Use of non intialized value. 1.7.2 ----- Returned scale estimates are computed using residuals at last iteration rather than n-1 iteration 1.7.1 ----- rcModelPLM, rcModelWPLM now accept an "input.scale" argument (and also return scale value) 1.5.3 ------ fix VECTOR_ELT/STRING_ELT issues 1.5.2 ----- normalize.quantiles.determine.target() and normalize.quantiles.use.target() now have a "subset" argument 1.5.1 ----- Default rcModelPLM() returns both probe coef and se for constraint row (previously SE was not properly returned) 1.3.6 ----- Fix issue with non double arguements to normalize.quantiles.use.target() 1.3.4 ----- Fix memory leak in determine_target Set pthread stack size where appropriate 1.3.3 ----- Fix memory leak in use_target 1.3.2 ----- R_subColSummary functions are now exposed at the C level 1.3.1 ----- rma.background.correct() was not correctly returning value when copy ==TRUE 1.1.9 ----- Commit missing c source files 1.1.8 ----- Commit missing header files 1.1.7 ----- Fix background function bindings (for use by other packages). The implementation themselves is unchanged. 1.1.6 ----- Addition of PLM-r and PLM-d Adjust rcModelPLM so that it takes optional row-effect estimates quantile normalization functions normalize.quantiles(), normalization.quantiles.determine.target(),normalize.quantiles.use.target() all now have multi-threaded support, user controlled using the R_THREADS environment variable Move weightedkerneldensity.c from affyPLM to preprocessCore subColSummarize* functions all now have basic multi-threaded support, user controlled using the R_THREADS rma background correction method (also multi-threaded) 1.1.5 ----- The subColSummarize* functions now return a matrix with rownames. Now it is clear which rows of the resulting summary matrix correspond to which values of the group.labels variable. This helps clarify the previous situation where it done alphabetically. 1.1.3 ----- Fix broken Makevars.in 1.1.2 ----- Add missing Makevars.in 1.1.1 ----- Add experimental support for pthreads based multi-threaded quantile normalization via code contributed by Paul Gordon <gordonp@ucalgary.ca> This is only implemented for the RMA quantile normalization (ie accessible via calling the rma() function. The number of threads is user controlled by setting the R_THREADS environment variable