The term ‘boot’ is now preferred for what used to be called ‘bootstrap’, mainly to avoid confusion with statistical bootstrap. To taf.boot() is similar to booting a computer, readying the components required for subsequent computations. Help pages now refer to ‘boot’, but all icesTAF functions fully support existing analyses that have a legacy ‘bootstrap’ folder.
The filename method.R is now an alternative to the default filename model.R script, for analyses where the term ‘model’ would be misleading or ambiguous.
See also: news(package=“TAF”, Version==“4.2.0”).
The ICES TAF R package has been split into two packages (TAF + icesTAF), mainly to facilitate the wider use of TAF beyond ICES. With the new design, the core functionality is in the TAF package, while icesTAF adds a thin layer on top of core TAF.
The core TAF package has no dependencies beyond base R, while the icesTAF package depends on TAF and possibly other packages.
Maintaining icesTAF as a light package will also facilitate further developments, e.g., tools that are specific to ICES workflows.
There is essentially no difference in the functionality of icesTAF 3.6.0 and icesTAF 4.0.0. Existing scripts will run as before and ICES TAF users continue to start their TAF scripts with library(icesTAF). This also loads and attaches the TAF package, so an explicit library(TAF) call is not necessary.
The only technical difference in the functionality of icesTAF 4.0.0 for the user is that if any existing scripts have used double-colon explicit namespace calls such as icesTAF::mkdir, then these should be changed to TAF::mkdir.
See also: news(package=“TAF”, Version==“4.0.0”).
add taf.boot.path() function to simplify access to boot folder
add function draft.data.script() to create R script template for downloading data with metadata
add local read.bib() function
fix taf2long so it works on tibbles
add taf.roxygenise() to build DATA.bib file from roxygen documentation
add function taf.sources() to extract data and software sources used in a project
move all bib entry processing to process.bib
add function taf.data.path() to provide path to bootstrap data
Added function clean.data() to selectively clean the bootstrap data folder.
Added function detach.packages() to detach all packages.
Added argument ‘overwrite’ to cp().
Added arguments ‘detach’ and ‘taf’ to sourceTAF().
Added argument ‘taf’ to taf.bootstrap().
Added argument ‘remove’ to taf.libPaths().
Added argument ‘details’ to taf.session().
Changed taf.bootstrap() to skip download of files if they already exist, and skip bootstrap script if the destination folder already contains files.
Changed taf.png() to use ‘res’ rather than ‘pointsize’, improving consistency between base, lattice, and ggplot2 plots. Code contributed by Iago Mosqueira.
Added function is.r.package() to check if tar.gz file is an R package.
Added functions long2xtab() and xtab2long() to convert between table formats.
Added function taf.libPaths() to add TAF library to search path.
Added function taf.session() to show session information.
Added argument ‘force’ to clean(), clean.library(), and clean.software().
Added argument ‘ignore’ to cp().
Improved clean.software() to remove software folder, not just software files.
Improved download.github() to store GitHub metadata in the DESCRIPTION file, if the GitHub resource is an R package. Warn if the tar.gz file looks like an R package nested inside a repository.
Improved taf.bootstrap() to support GitHub references in DATA.bib. Runs taf.install() only when software from GitHub contains a DESCRIPTION file. Added source = {folder} as a special value, similar to source = {file}. Warn if source = {owner/repo@ref} entry is missing the ‘@ref’ part.
Renamed internal function process.bib() to process.bibfile() and process.inner() to process.entry().
Improved handling of SOFTWARE.bib GitHub entries that have owner/repo/subdir. Code contributed by Ibrahim Umar.
Improved process.bib() so it verifies that ‘access’ values match the allowed values.
Added function get.remote.sha() to look up a SHA reference code on GitHub. Fixed a bug introduced in 3.2-0.
Improved clean.library() and clean.software() to remove the TAF library and software when SOFTWARE.bib file does not exist. Improved clean.software() to remove software file when version does not match SOFTWARE.bib.
Improved download.github() to handle packages that are nested inside a repository. Set default to quiet=FALSE.
Changed write.taf() so it no longer converts line endings.
Removed ‘remotes’ package dependency.
Added functions clean.library() and clean.software() to selectively clean the TAF library and software folders.
Added function download.github() to download a GitHub repository.
Added function taf.install() to install a package in the local TAF library.
Changed taf.bootstrap() so bootstrap/initial/config is now deprecated. Instead, model configuration files can be processed as a DATA.bib entry. Removed argument ‘config’. SOFTWARE.bib is now processed before DATA.bib.
Changed clean() so it uses clean.library() to clean the TAF library.
Improved draft.data() so it detects bootstrap/*.R scripts as source files to consider. Added arguments ‘access’ and ‘data.scripts’. Removed argument ‘data.dir’. Added support for file=TRUE as shorthand for file=“bootstrap/DATA.bib”.
Improved draft.software() so it first looks for packages in the TAF library. Added support for package=“owner/repo@ref” to install package on the fly. Added support for file=TRUE as shorthand for file=“bootstrap/SOFTWARE.bib”.
Improved taf.library() so it works from a bootstrap script.
Improved write.taf() so it gives an error if data frame contains comma.
Added argument ‘quiet’ to cp().
Added arguments ‘force’ and ‘recon’ to make().
Added argument ‘clean’ to process.bib(). Renamed argument ‘data.source’ to ‘data.scripts’.
Added argument ‘force’ to taf.bootstrap().
Improved convert.spaces() so it allows ‘sep’ to be any number of characters, treats %20 as space, and allows spaces in parent directories.
Improved download() so it shaves off ?tail from the end of ‘destfile’, and converts spaces and %20 to underscores.
Improved zoom() so it handles legend text size.
Removed ‘jsonlite’ package dependency.
Added function convert.spaces() to convert spaces in filenames.
Added function sam2taf() to convert SAM tables to TAF format.
Added argument ‘quiet’ to process.bib() and support for source={script} and dir={TRUE}. Removed ‘bundle’ field.
Added argument ‘quiet’ to taf.bootstrap(). Improved taf.bootstrap() so it doesn’t clean the corresponding bootstrap subdirectories when arguments ‘config’, ‘data’, or ‘software’ are FALSE. Dropped support for custom bootstrap.R script, superseded by source={script}.
Improved draft.data() so it works in R versions older than 3.5. Code contributed by Alexandros Kokkalis.
Changed sourceTAF() so working directory is not changed before script is run.
Added function period() to paste period string for DATA.bib entries.
Removed argument ‘name’ from taf.skeleton(). The default behavior is now to create initial directories and scripts in the current working directory.
Changed taf.library() so it loads a package instead of changing the library path.
Changed default size of taf.png(). Changed text size in zoom() accordingly and renamed arguments. The zoom() function is now generic.
Changed write.taf() so it gives a warning if column names are duplicated.
Improved process.bib() so it supports ‘prefix’ field when specifying multiple filenames in ‘source’ field.
Improved process.bib() so it doesn’t install a package that is already installed.
Added function sourceDir() to read all *.R files from a directory.
Added function zoom() to change text size in a lattice plot.
Added arguments ‘config’, ‘data’, and ‘software’ to taf.bootstrap().
Added argument ‘append’ to draft.data() and draft.software(). Code contributed by Alexandros Kokkalis.
Added argument ‘colname’ to xtab2taf().
Improved draft.data() so the default value of ‘year’ is the current year and user can pass period=FALSE.
Improved file.encoding() so it handles spaces in filenames.
Improved os.unix() so it recognizes both Linux and macOS operating systems.
Improved process.bib() so it supports ‘bundle’ field and multiple filenames in ‘source’ field, separated by newlines.
Improved write.taf() so file=“” prints to screen, regardless of dir.
Added function file.encoding() to examine file encoding. Added functions latin1.to.utf8() and utf8.to.latin1() to convert file encoding.
Added function line.endings() to examine line endings.
Added arguments ‘author’, ‘year’, and ‘title’ to draft.software(), and improved it so it also handles software other than R packages.
Added argument ‘clean’ to taf.bootstrap().
Added argument ‘create’ to taf.library().
Improved process.bib() so it does not require taf.bootstrap() to create ‘data’ and ‘software’ subdirectories.
Improved taf.skeleton() so every script starts with library(icesTAF) and mkdir() to create a working directory.
Added functions taf.bootstrap() and process.bib() to set up data files and software.
Added functions draft.data() and draft.software() to create initial draft versions of metadata files.
Added functions os(), os.linux() and os.macos() to detect more operating systems than before.
Updated colors (taf.green, taf.orange, taf.blue, taf.dark, taf.light) to make them equally intense and easy to distinguish.
This release introduces package dependencies to parse BibTeX files (bibtex) and install packages from GitHub (remotes), with further underlying package dependencies.
Added function taf.library() to work with packages in a local TAF library.
Added function rmdir() to remove an empty directory.
Added functions os.unix() and os.windows() to determine OS family.
Added function taf.unzip() to extract files from zip archives.
Moved functions read.dls() and write.dls() to the ‘icesAdvice’ package.
Changed clean(), makeAll(), makeTAF(), sourceAll(), and taf.skeleton() to align with new data-model workflow.
Renamed function tafpng() to taf.png().
Added argument ‘destfile’ to download().
Added argument ‘underscore’ to write.taf().
Added argument ‘stringsAsFactors’ to read.taf().
Improved multitable support for read.taf() and write.taf().
Changed sourceTAF(), sourceAll(), makeTAF(), and makeAll() so they never delete the ‘begin’ folder.
Changed sourceAll(), makeTAF(), makeAll() so they don’t set the working directory.
Changed clean() so user can specify directories to remove.
Changed write.taf() to search in the global workspace if ‘x’ is a string.
Changed cp() to enforce safeguards when moving files.
Removed function upload(). Initial data and model files are now in the ‘begin’ folder of each assessment.
Added function plus() to rename plus group column.
Added argument ‘dir’ to write.taf().
Changed read.taf() and write.taf() so they can read and write many tables in one call.
Changed write.taf() so the name of the data frame is the default filename.
Changed cp() to preserve the timestamp when copying a file.
Replaced argument ‘local’ with ‘rm’ in sourceTAF().
Added argument ‘colname’ to flr2taf().
Added functions makeTAF() and makeAll() to run TAF scripts as needed.
Added arguments ‘include’ and ‘engine’ to make().
Added argument ‘local’ to sourceTAF(), replacing the ‘rm’ argument.
Changed sourceAll() to only run TAF scripts: data.R, input.R, model.R, output.R, and report.R.
Added function upload() to upload file to TAF database. Added function download() to download file in binary mode.
Added function msg() to show a message, as well as the current time.
Added function tafpng() to open a graphics device. Added function lim() to compute axis limits.
Added function div() to divide column values with a common number. Added function rnd() to round column values.
Added function taf.skeleton() to create an empty template for a TAF analysis.
Added function deps() to list dependencies.
Renamed function sourceAtoZ() to sourceAll().
Added argument ‘move’ to cp(). Added argument ‘clean’ to sourceTAF(). Added argument ‘column’ to tt().
Added color objects: taf.green, taf.orange, taf.blue, taf.dark, and taf.light.
Added example data frame summary.taf to demonstrate div() and rnd().
Added function sourceTAF() to run scripts. Added cp() to copy files and mkdir() to create a directory. Added read.taf() and write.taf() to read and write TAF tables. Added long2taf(), taf2xtab(), tt(), and xtab2taf to convert between table formats.
Renamed functions readDLS() and writeDLS() to read.dls() and write.dls().
Removed function dir.remove() which is no longer needed in TAF scripts.
Added example data frames catage.long, catage.taf, and catage.xtab to demonstrate different table formats.