genomicSimulation v0.2.6

In brief:

Improvements to all input and output functions, to give users more control of the data they receive, as well as some bug fixes and under-the-hood improvements.

New Features

• Added ability to override the automatically detected file layout details in load.data and load.genotypes using the new function define.matrix.format.details, so that files whose format is incorrectly detected are not prevented from being loaded.
• see.group.data can now be called with multiple groups. The results from each group will be concatenated. This allows you to replace lines like c(see.group.data(group1,"BV"),see.group.data(group2,"BV")) with see.group.data(c(group1,group2),"BV").
• When accessing candidate names using see.group.data (i.e., for data.type = "N" or "P1" or "P2"), a missing name is no longer substituted with the candidate's pedigree ID. Instead, the name of a candidate with no name is rendered as an NA_character_. The old behaviour made it possible to accidentally pass string-typed pedigree IDs to make.targeted.crosses, and bugs could then result from make.targeted.crosses attempting to parse these as names.
• Added ability to directly observe the pedigree IDs of parents using see.group.data using data.type = "P1D" or "P2D".
• see.group.gene.data can now be called with multiple groups. The columns from each group will be concatenated.
• Column names and row names of the matrix returned by see.group.gene.data now reflect genotype names and genetic marker names respectively.
• Missing alleles in the matrix returned by see.group.gene.data are now visible (previously, they would cause issues with sometimes hiding non-missing alleles, due to being represented internally by the null character '\0').
• make.targeted.crosses now skips invalid pairings, instead of stopping execution at the first invalid pairing it finds. If any pairings were skipped, it will print a debug message listing the number of invalid pairings detected.
• The orientation of the output matrix of save.allele.counts can now be chosen, as has previously been possible only with save.genotypes.
• Save-as-you-go saving of breeding values calculated using a non-default set of marker effects is now possible. The save.gebv parameter of make.-prefixed functions now takes the identifier for a marker effect set, instead of a logical value. The default behaviour (not saving any breeding values) has not changed.

Improvements

• No longer Depends on package fs. fs is only used if file paths are used that contain tildes ~ needing to be expanded to home directories.
• Use of superseded progression function names (eg. cross.combinations for make.targeted.crosses) now produces warnings. The function names were superseded from v0.2.5 onwards.
• see.group.data and see.group.gene.data now raise a warning, instead of an error, when there are no group members allocated to the group(s) they access.
• see.group.data, see.group.gene.data, delete.group, and combine.groups no longer raise an error when the group numbers passed are of numeric type instead of integer type. Instead, as long as the numbers are whole numbers, they perform the type conversion. Likewise, make.targeted.crosses no longer raises an error when it is passed parent indexes that are of numeric type instead of integer type, as long as the numbers are whole numbers. Likewise, make.random.crosses, make.all.unidirectional.crosses, self.n.times and make.doubled.haploids no longer raise an error when the group numbers or map IDs passed are of numeric type instead of integer type, as long as the numbers are whole numbers.
• break.group.into.buckets now accepts integer bucket sizes even if they are not in R integer vector format.
• Functions which accept vectors of integers (eg. break.group.into.buckets, make.group, change.label.to.values) provide more meaningful error messages when passed a character vector instead.
• Vignette and function docs edited for better clarity around compatible input file formats and function name changes made in v0.2.5
• Under-the-hood improvements and simplifications of the genotype matrix input file layout detection systems.
• Under-the-hood improvements and simplifications of the file saving output functions.
• File saving output functions with multiple possible output formats (save.genotypes & save.pedigrees) now prefer a logical/boolean parameter to select the output format, the same way as the underlying C library does, instead of needing to memorise format strings. The old string parameters will still be accepted, so no need to modify old code.
• Changed name of startIndex parameter of change.label.to.values to skip to avoid confusion with genomicSimulation internal indexes (see.group.data(data.type="X")), which are not valid inputs to this function. The old parameter name is still accepted, so no need to modify old code.
• More informative error message when calling break.group.by.GEBV on a nonexistent group.
• Clarified in documentation which progression functions accept a vector of genetic map IDs, and which only accept a single genetic map.

Bug Fixes

• In load.data or load.genotypes, the number of markers per genotype that were successfully loaded is now accurately reflected in the printed log messages. Previously, these functions incorrectly printed out the total number of markers in the stored genetic map while claiming it was the number of markers in the genotype file that had been accurately matched to that map.
• Patched a couple of memory leaks in the underlying C library.
• Fixed a crash in make.clones with inherit.names = TRUE when genotypes being cloned had no names.
• Fix an infinite loop in make.random.crosses.between when both groups had breeding usage caps and one of the {caps x group sizes} was exactly the number of requested offspring.
• The give.names=TRUE setting in make.-prefixed functions no longer adds the unexpected prefix "NA" to names when no name.prefix is specified. If no name.prefix is specified, the names given will be identical to the pedigree IDs of the genotypes.