Changes in version 2025-05-05 (2025-05-05) - Remove includes of R_ext/PrtUtil.h from C code Changes in version 2017-05-22 (2017-05-22) - Small changes to avoid note about "R_registerRoutines" Changes in version 2015-07-23 (2015-07-23) - Rename plot.res as plot_res. - License to GPL-3 Changes in version 0.73-2 - Added NAMESPACE. Changes in version 0.72-1 - Fixed links in help files. - In anal.leaps, calls to summary.regsubsets() replaced with just summary(). Changes in version 0.71-7 - To get rid of a warning message in the latest version of R's package checks, changed the name of the functions convert.sim and convert.output to convertSim and convertOutput, respectively. Changes in version 0.71-6 - Fixed some slight problems in the help files. Changes in version 0.71-4 - Changed the name of the first argument in convert.sim and convert.output so that we don't get a warning message in R CMD check. Changes in version 0.71-3 - In calls to C code (using the .C() function), now make use of the PACKAGE argument. Changes in version 0.71-2 - The last column returned by which.correct() should be labelled "incor.unlinked" rather than "incor.total". This is now fixed. - Changed a call to error() to stop() in anal.leaps(). (There is no function error().) Changes in version 0.70-2 - Fixed a few minor problems in the help files. Changes in version 0.70 - Added forward selection following by backward elimination (method="forwback") to the function anal.leaps(). Changes in version 0.69 - Edited the documentation files to make them more complete (and to remove a repeated typographical error that resulted in the descriptions of the arguments not appearing in the help files). - Added a function convert.output() for converting the output of anal.leaps(), anal.mcmc() and perm() for use with which.correct(). - Fixed a bug in anal.leaps(): one line was hard-coded for exactly 100 markers! Changes in version 0.68 - Revised anal.multi2() back, so that MCMC starts at the null model. Changes in version 0.67 - Revised anal.multi2() so that MCMC starts at the model chosen by forward selection with BIC, for the first of the bic_mult given. Revised sim.mcmc() so that one may choose to do only forward selection and MCMC. Changes in version 0.66 - The C function forw_perm() was clearly wrong. I totally re-did it (thanks to advice from/discussion with Saunak Sen!) Changes in version 0.65 - There's a bug in anal.all() that causes a core dump. It turned out to be the simple problem of having the max.steps argument larger than the total number of markers. Changes in version 0.64 - Added anal.multi2() to do everything anal.multi() does, plus run MCMC with BIC-delta. Changes in version 0.63 - Added plot.cor() function for plotting the output of which.correct(). Revised which.correct() so that the last column is the number of unlinked (rather than total) extraneous loci. Changes in version 0.62 - Added functions combine.sims() combine.mcmc() to combine the simulation results of multiple runs of anal.multi() or sim.mcmc(), respectively. Changes in version 0.61 - Revised which.correct() to give both the total number of extraneous loci and the number of extraneous loci on the same chromosome as a QTL. Changes in version 0.60 - Added function to call the "leaps" packages to use the branch-and-bound procedure, and ported my old MCMC code to do a simple MCMC model selection procedure Changes in version 0.56 - Revised "anal.multi" to contain default LOD thresholds that I estimated via (extensive) computer simulations. Changes in version 0.55 - Added "anal.multi" function to do repeated simulations + analysis by ANOVA, CIM, and forward selection with BIC and permutation tests. Changes in version 0.54 - Completed addition of analysis programs, including the "bump finding" function necessary for turning LOD curves from ANOVA/IM and CIM into a selected model. Changes in version 0.53 - Added analysis programs. Changes in version 0.51 - Implemented an old (1996) simulation program. Why didn't I look for and reuse this code in the first place? Changes in version 0.50 - A new package; the aim is to test different model selection approaches to identifying QTLs in backcrosses, in the case of an additive model.