NEWS.md

In douglasgscofield/readGenalex: Read, Write, Manipulate and Convert GenAlEx-Format Genotype Files

readGenalex 1.0.9000

• Documentation tweaks

readGenalex 1.0

• Creates and manipulates data frames using the new class 'genalex'. The "genetic.data.format" attribute present in earlier versions of this package is removed.
• Extensive documentation updates reflecting use of class 'genalex'
• Several functions have been deprecated and replaced with generics recognising class 'genalex': printGenalexGenotype is replaced by printGenotype, computeGenalexColumns is replaced by getLocusColumns, reorderGenalexLoci is replaced by reorderLoci, getGenalexLocus is replaced by getLocus, putGenalexLocus is replaced by replaceLocus, dropGenalexLoci is replaced by dropLocus, and reduceGenalexPloidy is replaced by reducePloidy
• Added getPopulation to return genotypes for specific populations in a class 'genalex' object
• Extended is.genalex to have shallow (default) or deep (force = TRUE) checking of class 'genalex' consistency. With deep checking, attributes are inferred from the structure of the data frame and compared to the actual attributes. Checking of string attributes (dataset.title, data.file.name) and extra data columns may skipped with options.
• New as.genalex generic with as.genalex.genalex and as.genalex.data.frame. The latter will convert a pre-class 'genalex' data frame to class 'genalex', and is used wherever a data frame is returned so conversion from old to new format happens automatically. It will also convert a suitably-formatted data frame to class 'genalex' after determining suitable attributes
• There is a summary method for class 'genalex' which prints a few lines describing the data set, followed by a summary of the data frame contents and a summary of the extra columns, if present
• New as.data.frame method to convert class 'genalex' to class 'data.frame', optionally clearing all class 'genalex'-specific attributes
• New as.genetics method to convert class 'genalex' to a data frame with genotypes for each locus encoded using class 'genotype' or class 'haplotype' from package genetics
• New splitGenotypes function to split genotype-encoded genotypes into separate columns of a data frame
• New joinGenotypes function to join class 'genalex'-style allele columns into a single column per locus
• New as.loci.genalex method for the as.loci generic from package pegas to convert a class 'genalex' object to class 'loci', and as.genalex.loci for the reverse conversion
• New genalex function to create a class 'genalex' object from constituent data
• New rbind method to combine samples from multiple class 'genalex' objects
• New cbind method to combine loci and extra data columns from multiple class 'genalex' objects
• New addLocus method to directly add genotype data to the right side of an existing class 'genalex' object
• New extra and extra<- methods to get and set the extra columns of a class 'genalex' object. The latter coerces its value object to a data frame, if it is not already, does a bit of error checking, and sets the row names to be the sample names of the object.
• New ploidy method to get the ploidy of an object of class 'genalex'
• readGenalex now has ... in its argument list, for additional arguments to be passed to scan() while reading data
• Added na.strings argument to readGenalex, with defaults that reflect GenAlEx's own expectations for missing data
• Added na and na.character arguments to writeGenalex, with values to use for NA in genotype/numeric/logical columns and character columns, respectively
• Added readGenalexExcel and writeGenalexExcel which use the XLConnect package to read and write to Excel files of both .xls and .xlsx formats. These functions verify the consistency of their annotations and data prior to writing; this check can be removed on option.
• Added writeGenepop to write data in a class 'genalex' object to a Genepop-format text file
• Clarified typing of sample, population, genotype, and extra columns. Sample and population columns are stored as character (population was previously a factor), genotype columns are stored as numeric, and extra columns are stored as their natural type as determined by type.convert(..., as.is=TRUE), so that character columns are not converted to factors.
• As a result of the clarified typing, writeGenalex(..., quote=TRUE) will only quote names in the GenAlEx header, in the first two columns, and in any character-valued extra columns
• There is now more thorough checking for duplicate sample names, which should always produce an error
• There are also checks for duplication among the locus names and the sample and population columns
• The locus.columns attribute now has names which match the corresponding loci
• Added tests for all functions and found a few bugs along the way
• Rolled checkNullAlleles from my popgen repository into this package
• Extended checkNullAlleles to return a matrix of potential null alleles, with a couple of indexing modes
• Added quiet= option to checkNullAlleles, and pass through some arguments to printGenotype when quiet=FALSE
• Added Qagr_adult_genotypesand Qagr_pericarp_genotypes data sets and modified examples to use them, and removed the example_genotypes data set
• Created inst/extdata to hold data in non-R formats (simple tables), which of course includes GenAlEx-format data
• DESCRIPTION now has an explicit collation order and bug report URL
• Added 'data' target to Makefile to build specified RData files

readGenalex 0.4.1.9000 (development version)

• Moved documentation from separate *.Rd files to Roxygen2 comments in R source using the really helpful Rd2roxygen package.
• Implemented writeGenalex, to write a data.frame created by this package to a GenAlEx-compatible text file, and documented differences between output of writeGenalex and input to readGenalex
• Fixed bug in is.genalex when genetic.data.format attribute doesn't exist
• Slight expansion of documentation for reorderGenalexLoci
• Removed Date from DESCRIPTION, generated automatically during package creation
• DESCRIPTION uses Authors@R
• Marginally more clever Makefile

readGenalex 0.4.1

• First release available on CRAN
• First release available as a devtools-installable github repository (so development releases are always available via devtools::install_packages("douglasgscofield/readGenalex"))
• Restructured into R package directory structure, with individual exported functions split into separate files, and internal functions into readGenalex-internal.R
• Moved and expanded documentation following R package guidelines
• Added example_genotypes toy dataset. After loading the package with library(readGenalex) do data(example_genotypes) to load example_genotypes into your envirnment. The same dataset is also available here in text form as data/example_genotypes.txt.
• Added .Rbuildignore
• Improve error messages for sample size mismatches
• Handle lack of extra columns more gracefully, by completely leaving off the extra.columns attribute
• Fix getGenalexLocus() when given a pop= argument
• Fix printGenalexGenotype() to print sample name and population name using as.character() since population names are stored as factors
• Remove the data.column.names attribute, which duplicated the values in names

readGenalex 0.4

• Downconvert ploidy using reduceGenalexPloidy(). Currently only handles 2n to 1n, and selects only the first allele of each locus.

readGenalex 0.3

• Improve error messages when mismatch between group names in the header and in the data
• Handle the presence of extra columns to the right of the genotype columns. These are loaded into a separate extra.columns attribute.

douglasgscofield/readGenalex documentation built on May 15, 2019, 10:43 a.m.