NEWS.md
In steinbaugh/brio: Pipette Biological Data In and Out of R

Release notes

Major changes:

Updated dependencies to Bioconductor 3.18. These packages are now available on Bioconda.

Minor changes:

import: Reworked GAF file method to simply import as a 17 column data.frame. Refer to the Gene Ontology website for details on the GAF file format, which is basically a TSV files with commented lines beginning with "!".
Removed BaseSet as a suggested import. The getGAF function doesn't work on most of the GAF files available from the Gene Ontology website currently. It only works reliably with the Homo sapiens protein annotations, which isn't general enough for our needs currently.

Minor changes:

import: Added support for Gene Ontology (GO) annotation files (GAF). Useful for importing data from the Gene Ontology Annotation (GOA) database.
Miscellaneous documentation improvements for import.

Major changes:

export now supports list and S4Vectors List, for recursive export of supported objects. We may use this method in for other classes, such as SummarizedExperiment in AcidExperiment package in a future update.
export: GRangesList export method now inherits from List, and writes each object in list as a separate file to disk, rather than grouping in a data.frame coercion step first.

Minor changes:

export: GRangesList method is now defined and split out from data.frame method, to handle edge case of undesirable column name coercion.
Resolve new lints detected by lintr 3.1.1.

Major changes:

import now supports import of JSON and YAML lines via textConnection. This is very useful for importing return from CLI tools, such as aws.

Minor changes:

import: Now using read_yaml internally instead of yaml.load_file from the yaml package.

Minor changes:

import: Improved edge case handling of datasets with invalid names.
as.data.frame: Improved internal coercion code for Matrix and IRanges.
as.DataFrame: Ensure we set optional = TRUE internally during as.data.frame call.

New functions:

fillLines: Utility function for fixing malformed CSV and TSV files.
unfactorize: New generic function that intelligently converts a factor back to its original atomic data type.

Major changes:

The import and export generics are now defined in AcidGenerics rather than extending from BiocIO. This helps simplify the methods, removing options that we never use.
Now enforcing strict camel case for all function names.
The pipette file classes are now named in strict upper camel case.
Renamed cacheURL to cacheUrl.
Renamed getJSON to getJson.
Renamed getURLDirList to getUrlDirList.
Renamed pipetteTestsURL to pipetteTestsUrl.
Renamed removeNA to removeNa.
Renamed sanitizeNA to sanitizeNa.

Major changes:

getURLDirList: Entirely reworked internal code to no longer depend on RCurl package. Added support for HTTP(S) servers, which has been tested to work for Ensembl and NCBI.
import: Now supports textConnection class for primary con argument instead of always using character representing a file path. This is incredibly useful for reformatting a malformed delimited file on a remote server prior to import, which can help eliminate our reliance on readr or data.table as alternative engines to handle malformed files.
transmit: Reworked internal code to no longer depend on RCurl.

Minor changes:

export: Now supports GRangesList, which first coerces to data.frame. This class includes "group" and "groupName" as the first columns upon export.

getURLDirList: Tighten up assert checks to intentionally error if input does not contain an FTP server URL.

Major changes:

After inspecting the package with packageDependencies from AcidDevTools, decided to revert the changes introduced in 0.12.0 back to making more heavy dependencies optional. We are now no longer requiring BiocFileCache, data.table, digest, httr2, readr, and rtracklayer to be installed. This helps speed up attachment of the package, where it is most commonly used to simply import and export CSV files as the primary utility.
These package changes will not affect the bioconda recipe, as all optional dependencies are bundled with the recipe.

Minor changes:

getURLDirList: Need to move isAnExistingURL assert check down, to not break Ensembl FTP server checks in AcidGenomes makeGRangesFromEnsembl.

Minor changes:

as.DataFrame: Fixed support for nesting of DFrame objects.
droplevels2: Ensure that original class and object metadata are preserved.

Major changes:

Migrated a number of dependency packages from Suggests to Imports (see DESCRIPTION and IMPORTS files) to avoid issues with missing dependencies in some commonly used Acid Genomics software. Here we are now requiring BiocFileCache, data.table, digest, httr2, jsonlite, readr, rtracklayer, and yaml as standard packages in pipette, so that they get automatically installed. This change only applies when managing dependencies using R directly and doesn't affect bioconda recipes.
Hardened some internal assert checks using isAnExistingURL instead of simply isAURL. This new function available in the goalie package actively checks to see if the URL exists and is active. Note that for FTP directories, such as with the transmit function, the isAnExistingURL check only works if there is a trailing slash.

Minor changes:

Disabled examples using the Ensembl REST API server ("rest.ensembl.org"), as this has recently been flaky and can cause build checks to time out.
Miscellaneous unit test fixes, now using isAnExistingURL to check.

Minor changes:

import: Disabled lazy loading mode for readr. Enabling this makes it basically impossible to suppress warnings when parsing malformed files, such as in the MGI function of AcidGenomes.
Enabled parallel testthat unit tests.

Minor changes:

getJSON: Improved code coverage for parsing of Ensembl REST API.

Major changes:

Now requiring R 4.3 / Bioconductor 3.17.
File classes for import and export are now prefixed with Pipette, to avoid unwanted collisions with other classes defined in Bioconductor packages.
import: Added support for MAF files.
import: Added support for BAM, CRAM, and SAM files. Note that CRAM files must be able to resolve the corresponding reference genome.
import: Added support for BCF and VCF files. Note that these files currently require corresponding CSI index files, which can be generated using htslib.
Temporary files generated during import calls (i.e. automatic decompression of compressed files, such as gz or xz) are now automatically cleaned up. This change should only affect temporary files. If you notice any issues with this, please file a bug report!
getJSON: Reworked to use httr2 instead of deprecated httr package.

Major changes:

import: Improved support for FASTA files. The function will now intentionally error on any warnings, which can occur if the moleculeType argument is not set correctly. For example, we're defaulting to DNA input here, but that is not always the case. miRBase FASTA files are in RNA format, so use moleculeType argument here to set "RNA" instead of "DNA". We've also added amino acid support via "AA" argument, which passes to Biostrings package in a similar method for DNA and RNA. Note that for miRBase FASTA files, we now attempt to get metadata from the FASTA identifiers, which are defined as a DataFrame slotted into metadata as attributes.

Minor changes:

import: Updated data.table engine settings to use header = "auto" when colnames argument is declared as TRUE. This will attempt to handle malformed columns, similar to readr/vroom approach.
Improved internal usage consistency of goalie boolean grepl wrapper functions, such as isMatchingFixed and isMatchingRegex, which improves code legibility.
Removed legacy Docker usage instructions in README.

Minor changes:

Renamed S4 class definitions: DataFrame to DFrame; GenomicRanges to GRanges; and IntegerRanges to IRanges.

Minor changes:

import and export now support overriding default quote handling with the quote argument. Changing this is not recommended by default, but is useful for some edge cases with annoying gene metadata files from Ensembl. Who decided that B" is an acceptable gene name?

Minor changes:

import: Switched default quoting for delim file using base engine. Also removed empty string assert checks, which can be problematic for some data types returned by readr and data.table engines.

Minor changes:

Added line break to naStrings handling, which improves automatic sanitization of some malformed gene CSV files on the Ensembl FTP server.

Minor changes:

Reexporting initDir and pasteURL from AcidBase.
Improved pattern documentation for getURLDirList.

Minor changes:

as.DataFrame: Reworked the list and SimpleList methods:
We removed the option to set row.names for list and SimpleList methods. Instead, row names will get defined automatically when the first element in the list is named.
The list method now checks for length mismatched input at the start of the function. This should also be handled in the downstream DataFrame generator step, but we can set a custom error message instead.
We're now calling new internally with listData and nrows defined, which is faster and less problematic then attempting to construct a matrix (via cbind) and then coerce to DataFrame instead.
Our list method now attempts to reorder named elements when possible.
decode: The internal code for DataFrame method has been simplified, due to rework of our internal as.DataFrame method code. Note that list elements are now unname during the decoding step, but has no effect on row name handling.
droplevels2: Hardened DataFrame method to properly coerce DataFrame to SimpleList before proceeding. Using as method with "List" class does not currently unclass DataFrame as expected.

Minor changes:

Migrated requireNamespaces import in NAMESPACE from AcidBase to goalie.
Updated package dependencies.

Minor changes:

factorize: Bug fix to ensure that logical columns are not coerced to factor, as these are expected to have repeated values.
Now requiring Bioconductor 3.16 release.

Major changes:

import: Added support for import of gene cluster text (GCT) file format. Currently can set return parameter to import as a matrix (default) or data.frame with Name and Description columns retained.

Major changes:

This release attempts to harden the primary arguments supported for export and import. Note that file is intentionally not supported in favor of con, which is the preferred convention used in BiocIO.
Reduced the number of exported methods, tightening on missing instead of using missingOrNULL, which can cause unwanted inheritance issues that collide with BiocIO package.
export now supports any type of atomic vector.

Minor changes:

factorize: Improved handling of DataFrame input that contains complex S4 columns, such as CharacterList.

Major changes:

factorize: Reworked internal method to only coerce columns with repeated values to factor. Now uses anyDuplicated internally to check for which columns to factorize.

Minor changes:

import: Hardened the internal readr engine to use base make.names for column name repair for delimited file import (e.g. CSV, TSV). This also eliminates unwanted CLI messages about name repair when using readr.

Minor changes:

export: Bug fix to provide compatibility in Matrix method for export of sparse matrices without dimnames (e.g. row and/or column names). Hit this edge case in the update of example data in our Chromium package.

Minor changes:

Updated and hardened unit tests to avoid file system lock issues on Windows related to readr engine.

Minor changes:

atomize: Hardened edge cases of empty DataFrame and GRanges input. Also improved code coverage to test handling of these edge case events, which can occur when exporting metadata from SummarizedExperiment in upstream AcidExperiment and bcbioRNASeq packages for objects with minimal metadata.

Minor changes:

Updated lintr checks and testthat unit tests.

Minor changes:

import: Added support for Open Biomedical Ontologies (OBO) format. This uses ontologyIndex internally. See also BiocSet for alterative import method that uses es_set and other accessor functions.

Major changes:

Switched primary import/export engine from readr back to base, to avoid strong dependency on readr package. Note that both readr and/or data.table packages can optionally be used for import/export by specifying "engine".
Removed coercion method support for as_tibble and as.data.table, to remove strong dependencies on data.table and tibble packages.
Removed as coercion support in favor of simply using as.DataFrame S4 generic approach, to avoid potential conflicts with Bioconductor.
as.DataFrame: Reduced the number of supported classes, removing support of non-Bioconductor classes, specifically data.table and tbl_df.
Removed re-exported functions: column_to_rownames, getURL, rbindlist, rownames_to_columns, and tibble.

Minor changes:

Package code is now formatting using styler package.
Now exporting method support in droplevels2 instead of droplevels, to avoid method conflict with new Bioconductor 3.15 update.
All requireNamespaces calls are now wrapped by assert.
Removed previously deprecated functions that are no longer in use. These were previously defined in deprecated.R: sanitizeColData, sanitizeRowData, sanitizeRowRanges, and writeCounts.

Major changes:

No longer reexporting any functions or S4 classes.
export, import: Switched to recommended new BiocIO generic approaches, which are now used in multiple Bioconductor packages, notably rtracklayer.
export, import: Switched back to readr as default engine from data.table. The data.table fread and fwrite functions have been shown to generate stack imbalances in some edge cases that readr handles better.
export: Current default recommended method now dispatches on "object", "con", and "format" arguments. Previous methods that dispatch using "format", "dir", "ext", and/or "file" (adapted from the conventions used in rio package) arguments are soft deprecated, but should still currently work. If you encounter any breaking changes here, please file an issue!
import: Current default recommended method now dispatches on "con", "format", and "text" arguments. We have defined our methods to simply handle the "con" argument as a file path. Note that the "text" argument can be useful for passing in raw lines of a particular file format, but we are not currently supporting that edge case in any of our import methods yet. Such functionality may be added in a future update.
export data.frame / DataFrame methods now attempt to coerce nested list columns to vector via toString internally when possible.

Minor changes:

import: Added rownameCol argument for import of delimited files, such as CSV, TSV, and Excel. This is a non-breaking change that enables the user to manually define the rowname column upon import, which can be useful when working with files from public databases such as GEO.
Updated suggested readr and vroom versions, now that they're available on Bioconda.

Major changes:

Switched bapply import from AcidBase to goalie.
Updated internal basejump dependencies.
Improved installation instructions.

Major changes:

Updated minimum Bioconductor release to 3.13.
import: Added support for FASTA and FASTQ files, which are loaded via Biostrings package internally. Refer to readDNAStringSet for details.
import: Improved makeNames consistency for methods that import two-dimension arrays (e.g. data.frame).

Minor changes:

Improved package error messages and other alerts with AcidCLI update. Instead of calling stop internally, now using abort, which supports stylized messages (via cli package).
transmit: Hardened working example against NCBI FTP server failure.
Package is now back to 100% code coverage, with improved coverage of primary import and export functions.

Minor changes:

sanitizeNA(): Reworked internal call to use factor instead of as.factor followed by a separate levels call. This can result in an unwanted value swap for factors with a single value. Added code coverage to check for this.
Updated package dependency versions.

Minor changes:

localOrRemoteFile: Improved handling of URLs without a file extension.
import/export: Simplified the appearance of file variables, making them easier to copy directly from the console for debugging.

Minor changes:

import: Ensure that data.table engine always interprets empty strings ("") as NA. The fread function is opinionated about this and doesn't currently respect "" input as an NA string.
decode/encode: Ensure we're not dropping metadata here.

Major changes:

import now uses S4 methods based on the file argument, which can be manually overridden with the format argument (e.g. "csv" for a CSV file).
import and export now support engine argument for methods dispatching on character and data.frame. Currently "base" (base R), "data.table", "readr", and "vroom" are supported.
Source code lines now default to using base R for import/export by default. Previously, this defaulted to data.table, which is now only used as the default for import of delimited files.

Minor changes:

import: Added support for removeBlank and stripWhitespace for import of source code lines (LinesFile).
import: Improved import handling of source code lines using data.table.

Minor changes:

decode / encode: Reworked internal code slightly to provide compatibility with S4Vectors update in Bioconductor 3.13.

Major changes:

as.DataFrame: Updated list to DataFrame coercion support that is compatible with Bioconductor 3.13 release update.
cacheURL: Now using tools::R_user_dir instead of rappdirs::user_cache_dir internally. This matches the conventions used in Bioconductor 3.13 (e.g. AnnotationHub and BiocFileCache).

Minor changes:

Documentation updates, to pass build checks without warnings on R 4.1 and Bioconductor 3.13.

Minor changes:

cacheURL now internally calls BiocFileCache and rappdirs as suggested packages, rather than direct imports. This helps keep the package a bit lighter and improve loading times, as BiocFileCache currently calls a number of heavy dependencies, including dplyr.
The checksum functions md5 and sha256 now call digest internally as a suggested package, rather than a direct import.

Major changes:

Switched the default engine back to data.table package for import and export functions.

Minor changes:

Improved consistency of arguments for internal engines (base, data.table, readr, and vroom) for import and export of delimited files.
Base engine now uses read.table and write.table instead of read.csv and write.csv for CSV files.
The optional readr engine now uses read_delim and write_delim, similar to the update for base engine.
Improved code coverage of base engine for import/export.

Minor changes:

import: Hardened importer against unexpected mismatch when user attempts to manually define column names via colnames argument. Some importers such as vroom are currently too liberal about mismatches.

Minor changes:

Including seqnames as a reexport, which is defined in GenomeInfoDb via GenomicRanges.

Minor changes:

as.DataFrame: Restrict list-based coercion to list and SimpleList (instead of List virtual class).

Minor changes:

Include CompressedGRangesList from GenomicRanges as a reexport.

Minor changes:

Migrated IRanges reexports to AcidGenerics.

Minor changes:

Updated naStrings to include "-", "_", and " ".

Minor changes:

Minor rework and simplification of NAMESPACE, inheriting from AcidGenerics and AcidBase when possible.

Minor changes:

Reexporting additional useful functions and classes from IRanges, including AtomicList virtual class.
Migrated S4Vectors reexports to AcidGenerics, including Annotated, Factor, Factor, LLint, and RectangularData classes.

Minor changes:

Including more reexports of useful S4 classes and functions defined in IRanges that we will reexport in basejump: CharacterList, FactorList, IntegerList, LogicalList, NumericList, RleList.

Minor changes:

Reexporting some additional functions and classes from GenomicRanges, IRanges, Matrix, and S4Vectors that we can inherit in basejump.

New functions:

Added md5 and sha256 functions, that use the digest package internally. Previously these were defined in the AcidGenomes package, but have migrated here for file management consistency.

Major changes:

cacheURL: Fix bug that resulted in different remote URLs with the same base name being cached as the same object internally by BiocFileCache. Now URLs should always be cached uniquely.

Minor changes:

Migrated test data from "tests.acidgenomics.com" to "r.acidgenomics.com/testdata".
Added droplevels method for DataFrame.

Minor changes:

cacheURL: Improved message. Now only showing when file gets cached into package cache via BiocFileCache. Now defaults to caching into BiocFileCache directory instead of pipette.
Now exporting useful rbindlist function from data.table.
Renamed matchRowNameColumn to matchRownameColumn (note case).

Minor changes:

naStrings: Now including lowercase NA variants, which are seen in some files on RefSeq FTP server.
sanitizeNA: Updated to also match lowercase NA patterns.

Major changes:

import: Improved internal engine support for plain text delimited (e.g. CSV, TSV) files and source code lines. The vroom engine remains enabled by default but data.table, readr, and base R are consistently supported for import of either delimited files or source code lines.
export: Added supported for base R export of CSV and TSV files. Internal engine consistency has been improved for character and matrix/data.frame methods. Note that character method currently falls back to using readr's write_lines function instead of attempting to use the vroom package by default.

Minor changes:

getURLDirList now returns sorted.
Reexporting url.exists function from RCurl, for convenience.

Minor changes:

Switched from using cli package internally to AcidCLI.

Minor changes:

Made some previous imports conditional suggested packages: jsonlite, readr, rtracklayer, yaml.
export: Improved internal file name handling for CLI messages.
Reworked rtracklayer as a suggested package instead of an explicit import.
import: Improved internal bcbio counts importer code to use default TSV method, rather than relying on data.table fread function manually.
Removed dependency on readr. Import of lines now uses vroom::vroom_lines internally, and export character method will conditionally switch to using base writeLines if the readr package is not installed.
Removed Matrix readMM and writeMM as imports.
Removed data.table fread and fwrite as imports.
Removed BiocGenerics dependency, in favor of rexports defined in AcidGenerics. This helps keep the number of dependencies declared in the package more compact and manageable.
export: Bug fix for handling of GZ file name extension for character method.
Bug fix for vroom_lines error import error of bcbio log: Unnamed col_types must have the same length as col_names.
Now including additional reexports from data.table and tibble packages.
Bug fixes for import/export of tx2gene file handling for pending AcidGenomes package update.
import: Hardened against makeNames usage on objects that don't support names assignment.

Minor changes:

export: Added option to intentionally not export column and/or rownames for matrix, data.frame, and DataFrame classes.

New functions:

Added download, which acts as a hardened wrapper for utils::download.file. Annoying, download.file returns status codes but does not intentionally error on any unsuccessful downloads. Our wrapper ensures that R always errors on any file download issue. It also sets a longer timeout internally, to avoid any potential issues with the timeout option being defined in Rprofile.

Minor changes:

Updated dependency versions.
export: Added append option for character method. Also relaxed checks on character method, allowing for exporting of empty vectors.

Minor changes:

transmit: Added download argument support, to optionally return matching URLs without downloading. This is useful for handing off to cacheURL function for caching files inside of packages with BiocFileCache.

Minor changes:

Bug fix for breaking change in readr v1.4 release. In the write_* functions, including write_lines and write_csv, the path argument has been renamed to file. Now requiring readr v1.4+ in pipette.

New functions:

Migrated transmit here from basejump.

New functions:

getURLDirList: Migrated this function from previous definition in basejump, so we can inherit inside new AcidGenomes package.

Minor changes:

cacheURL: Added package argument, so other packages that use this function will automatically inherit the current package, as expected.

New functions:

cacheURL: Utility function for easy package file caching using BiocFileCache package internally.

Minor changes:

Updated Acid Genomics package dependencies.

Minor changes:

sanitizeNA: Added support for "N/A" string, which is present in some Excel spreadsheets.

Minor changes:

export: Now dropping non-atomic columns (e.g. Entrez ID list column) from data frames automatically prior to export. Previously, the allAreAtomic assert check was called automatically and would error on non-atomics.

Minor changes:

import: Improved messages to always resolve full path to import directory.
Bug fix for AppVeyor CI config.

Minor changes:

export: Ensuring that full directory path is always resolved in message.
Miscellaneous message improvements, related to internal toString handling.

Minor changes:

Relax name checks for import.

Minor changes:

Decreased data.table dependency from 1.13.0 back to 1.12.8, so we can build successfully on bioconda.

Minor changes:

export: Improved messaages to include full output path.
Increased minimum R dependency to 4.0.

Minor changes:

import Hardened Excel input to intentionally error on any warnings returned by internal read_excel call, which is too liberal in coercing data types, in my opinion.
import: Switched XLS parser from gdata (no longer updated) back to readxl, which is more actively developed.
import: Added makeNames argument, to override default internal handling. This allows the user to apply snake case and/or camel case formatting automatically with this argument.

Minor changes:

import: Added support for skip argument, which allows the user to skip a certain number of lines in the input.
import and export of source code lines now uses readr package internally (read_lines and write_lines) instead of base readLines and writeLines.

Minor changes:

import: Now setting delim internally for vroom import call, to handle single column data frame import. Otherwise vroom will warn about failing to detect expected delimiter.

Major changes:

import and export functions now default to using vroom engine instead of data.table. Internally, these now call vroom and vroom_write. We noticed that the data.table fwrite function in particular can have issues writing many files on AWS EC2 instances, resulting in a stack imbalance. The vroom package seems to be more stable currently.

Minor changes:

loadData, saveData: Switched to using cli_alert instead of cli_text internally for status messages.

Minor changes:

import: Bug fix for format argument erroring on some supported file types.

Minor changes:

droplevels: Ensuring S4 generic variant defined in S4Vectors package gets reexported and masks base S3 generic. This helps avoid a C stack usage issue that has popped up in the latest version of R.
import: Fix for importing JSON files without extension. Can now declare using the format argument. This fix was needed to import GitHub JSON URLs inside new installGitHub function defined in bb8 package.

Minor changes:

Switched license from MIT to GPL-3.
Ensure that coerce method is reexported -- thanks @dpryan79 for catching this issue in basejump.

Major changes:

Renamed package from brio to pipette, in preparation of CRAN submission.
export: Reworked internal methods to use new compress and decompress functions defined in acidbase package.
localOrRemoteFile: Reworked to use new decompress defined in acidbase internally, with improved tempfile handling.
Migrated coercion methods and other utitilies from the now archived transformer package: atomize, coerceToList, droplevels, decode, encode, factorize, matchRowNameColumn, and metadata2.

Minor changes:

Switched to using cli package for improved messages.

Minor changes:

import: Added metadata parameter option and improved internal tryCatch handling if call capture fails. This can be the case when nesting the function inside another function, which can cause standardizeCall to fail. Note that match.call doesn't have this problem but doesn't consistently expand the call with default formals as well.

Minor changes:

export: Removed internal dependencies on as_tibble and as.data.table calls. The tibble package recently changed the default row name handling behavior in as_tibble, which broke the code here. I reworked the internal code to only use base R approaches, so changes in the tidyverse no longer affect the package.

Minor changes:

Updated package dependencies to require Bioconductor 3.10 release.

Minor changes:

Improved internal metadata handling using new metadata2 function.

Minor changes:

NAMESPACE updates to support migration of some low-level functions to the new acidbase package.

New functions:

getURLDirList: Return a simple character vector of files and subdirectories in a remote directory. Intended for use with FTP servers.
Also now reexporting the getURL function from RCurl.

Minor changes:

import: Default format argument has been renamed from "none" to "auto".
localOrRemoteFile: Improved handling for remote URLs without a file extension.
pasteURL: Now smartly strips trailing slashes prior to internal paste call.

naStrings: Reverted back to including only "NA" and "NULL", instead of including empty space strings. This results in unwanted messages regarding strip whitespace from data.table fread function.

Major changes:

Added back internal support for readr package instead of data.table for import and export functions. We have observed stack imbalance and segfault memory dump issues with the latest data.table release (v1.12.4) on multi-core Azure VMs. The engine can now be changed using global options:
import: acid.import.engine ("data.table" or "readr").
export: acid.export.engine ("data.table" or "readr").

This new addition is experimental and may be dropped in a future release. We find that readr currently works more reliably for export in some cases for large CSV files, but data.table is generally faster and more robust for data import of CSV and TSV files. We're intentionally keeping this functions simple and not providing a user-facing argument for selecting the internal engine.

Minor changes:

Converted unnecessary exported global variables into internal globals: compressExtPattern, extPattern, rdataExtPattern, rdataLoadError.
Updated naStrings to include empty whitespace.

Minor changes:

Updated data.table and rio dependencies, based on recent data.table 1.12.4 update, which has a lot of changes.
Fixed internal code to no longer show rownames message, even when rownames = FALSE.

Major changes:

import: Improved metadata stash approach inside S4 (metadata) and S3 (attributes) return objects. Previously, import metadata was stashed inside "brio", but this has been renamed to "import". Metadata is no longer stashed inside R data objects loaded via import. Simplified the internal code inside import to only stash the call, whereas importer metadata is now handled by an internal .defineImportMetadata function.

Minor changes:

export: Improved default extension documentation and internal argument matching via match.arg. For matrix files exported simply with dir argument, this will default to CSV format. For sparse Matrix files exported simply with dir argument, this will default to MTX format.

Minor changes:

import: Improved internal arguments passed to data.table::fread.

Updated R dependency to 3.6.

Major changes:

export: Reworked internal code to call data.table::fwrite directly, rather than having to pass to brio::export. Added support for bz2 output for matrix and sparseMatrix classes.

Minor changes:

No longer warn on MTX import without sidecar rownames and colnames files.
import: Don't attempt to slot attributes for atomic vector return. This applies to source code lines and helps avoid valid name issues when assigning these values to colnames or rownames. I came across this issue while updating the Chromium package to assign names from 10X Genomics sidecar files.

Minor changes:

Reworked organization and naming of internal importer functions. Now including these in the import documentation, for clarity.
import: Added format and setclass arguments.
Switched to data.table::fread for import of bcbio count matrix. Previously, was using readr::read_tsv.
Import of lines no longer stores brio attributes.
Reduced number of package dependencies, no longer requiring readr.

Minor changes:

Added support for url calls to import, localOrRemoteFile, and loadRemoteData.
Improved message consistency.
Updated basejump dependency versions.

Minor changes:

Now using acidroxygen package to manage shared roxygen documentation params.

Minor changes:

import: Bug fix for invalid objects (e.g. S4 objects that inherit SummarizedExperiment) not returning silently. Had to convert the try call to a tryCatch call to avoid errors popping up during name checks.
import: Now suppressing partial match of 'OS' to 'OS.type' warning for import of XLS files, which is a bug in gdata package.

Bumped version number to reflect changes in basejump dependencies.

Minor changes:

Improved naming consistency of internal functions.
Updated package dependency versions.

Minor changes:

Improvements to Travis Docker and AppVeyor CI checks.

Minor changes:

factorize: Tightened up method support. Now exporting DataFrame only.
Bug fix for acid.data.frame global option support.
sanitizeNA: Improved factor return.
Improved code coverage.

Major changes:

loadData, loadDataAsName, and loadRemoteData now support overwrite argument. The default behavior of these functions has changed to allow overwriting into the environment by default, matching the base load function conventions. If this behavior is undesired, set options(acid.overwrite = FALSE) and this will be inherited in all calls.
Renamed acid.export.overwrite and acid.save.overwrite to simply use acid.overwrite global for IO functions. This was modified now that loadData also supports the overwrite argument.

Major changes:

import: Removed Google Sheets support. The googlesheets R package is currently too buggy, and the replacement googlesheets4 package isn't stable. This functionality may be added back in a future update.

Introducing breaking changes to export method. Now using object instead of x as the primary argument, and defaulting to the use of ext and dir instead of recommending the use of file, as is the convention in the brio package. This makes interactive file export quicker and more intuitive, involving less repetitive variable declarations.

Minor changes:

S4 generic reexport documentation fixes.

Minor changes:

Backward compatibility fixes/updates to support R 3.4.

Minor changes:

Bug fix release. Re-importing rio package to ensure export() always works on data.frame method.

Minor changes:

import: Added GRP file support for GSEA.
import: Improved Google Sheet import support, using new googlesheets4 package from tidyverse.
Improved code coverage, getting closer to 100%.

New functions:

Migrated sanitizePercent here from basejump package.

Major changes:

Switched to using "acid" prefix instead of "basejump" for global options. This applies in particular to the loadData and saveData functions, where the dir argument can be set globally for an interactive session using this parameter.

New functions:

Migrated removeNA and sanitizeNA from basejump here, so these functions can be imported in freerange package.

Minor changes:

Migrated code to Acid Genomics.

Minor changes:

localOrRemoteFile: Improved error message for Windows users when tempfile can't be removed successfully. This can happen on systems when the user is not running as Administrator, but doesn't happen on macOS or Linux.

Minor changes:

import: Added initial rownames and colnames parameter support for data.frame import. I strongly recommend leaving these enabled by default. However, these are useful in some edge cases when loading data from remote servers (e.g. WormBase).
export: Removed ... passthrough for data.frame method.

Minor changes:

import: Switched to using gdata::read.xls to import legacy XLS binary files. readxl::read_excel doesn't work consistently for some files and returns libxls error. In the meantime, use gdata package, which is slow but works.
Removed validObject check in import call.

Minor changes:

localOrRemoteFile: Binary file extension pattern matching bug fix. Applies to files downloaded on Windows. If download.file is not called with mode wb (write binary) for files on Windows, decompression will fail.

Major changes:

loadData and saveData now support list argument, supporting a character vector of object names. This allows for programmatic use of the functions with standard evaluation. For reference, this approach is inspired by the method defined in save.

Minor changes:

import: Data provenance metadata is now slotted into attributes for S3 return (e.g. data.frame) and metadata for S4 return (e.g. DataFrame).

Minor changes:

export: Tightened up the method support, removing the ANY method. Now we're explicitly exporting data.frame, DataFrame, matrix, and GRanges methods. This also has the added benefit of making the documentation more readable.

New functions:

fileExt: An improved variation on tools::file_ext.

Minor changes:

basenameSansExt: Tightened up this function to return NA on match failure. This behaves simiarly to the new fileExt function.
Switching back from defunct to deprecated for sanitizeColData, sanitizeRowData, sanitizeRowRanges, since these functions are still in use by bcbioRNASeq v0.2.9.
Improved NEWS file for previous releases.

Minor changes:

Documentation fixes and website improvements.

This release helps ensure backward compatibility with R 3.4.

Minor changes:

Bug fix for assert in transmit: Need to wrap isMatchingRegex in all for backward compatibility with R 3.4.
Miscellaneous CI fixes to Travis CI and AppVeyor CI.

New functions:

These functions have been migrated from basejump here to brio, since they deal specifically with file input/output:

atomize.
decode.
encode.
factorize.

Removed functions:

Removed sanitizeColData and sanitizeRowData. May need to add back in a future release, but removed for time being.

Minor changes:

realpath: Removed unnecessary assert check using allHaveAccess, since normalizePath already checks for this when mustWork = TRUE.
Added initial code coverage support using testthat.

Minor changes:

Updated imports to reflect renaming of S4Transformer package to simply transformer.

Minor changes:

Improved extPattern to inherit compressExtPattern.
Improved documentation in import regarding compressed file handling.
Split out NAMESPACE imports into imports.R file.
Updated sanitizeColData and sanitizeRowData to take advantage of exported atomize function.
Improved Travis CI and AppVeyor CI configuration.

Minor changes:

Added Travis CI and AppVeyor CI support.
Improved documention in import regarding supported file formats.
Split out internal importers into separate R files.
Fixed return value for pasteURL.
Disabled working examples for transmit, since they are failing on Travis CI.
Documentation fixes and miscellaneous tweaks to pass build checks.

Initial release. Migrated input-output (IO) functions from basejump.

steinbaugh/brio documentation built on Feb. 1, 2024, 6:42 p.m.

rdrr.io home R language documentation Run R code online

CRAN packages Bioconductor packages R-Forge packages GitHub packages

Note that we can't provide technical support on individual packages. You should contact the package authors for that.

steinbaugh/brio Pipette Biological Data In and Out of R

NEWS.md In steinbaugh/brio: Pipette Biological Data In and Out of R

Release notes

pipette 0.15.2 (2023-12-15)

pipette 0.15.1 (2023-12-13)

pipette 0.15.0 (2023-11-09)

pipette 0.14.2 (2023-11-03)

pipette 0.14.1 (2023-10-06)

pipette 0.14.0 (2023-10-03)

pipette 0.13.0 (2023-09-19)

pipette 0.12.4 (2023-09-15)

pipette 0.12.3 (2023-09-14)

pipette 0.12.2 (2023-09-13)

pipette 0.12.1 (2023-09-12)

pipette 0.12.0 (2023-08-25)

pipette 0.11.2 (2023-08-12)

pipette 0.11.1 (2023-07-27)

pipette 0.11.0 (2023-07-13)

pipette 0.10.10 (2023-06-29)

pipette 0.10.9 (2023-04-26)

pipette 0.10.8 (2023-04-13)

pipette 0.10.7 (2023-04-13)

pipette 0.10.6 (2023-04-13)

pipette 0.10.5 (2023-04-12)

pipette 0.10.4 (2023-02-22)

pipette 0.10.3 (2023-02-09)

pipette 0.10.2 (2023-01-31)

pipette 0.10.1 (2022-11-10)

pipette 0.10.0 (2022-10-25)

pipette 0.9.7 (2022-08-19)

pipette 0.9.6 (2022-08-19)

pipette 0.9.5 (2022-06-07)

pipette 0.9.4 (2022-06-02)

pipette 0.9.3 (2022-05-25)

pipette 0.9.2 (2022-05-23)

pipette 0.9.1 (2022-05-12)

pipette 0.9.0 (2022-05-04)

pipette 0.8.0 (2022-03-11)

pipette 0.7.2 (2021-09-22)

pipette 0.7.1 (2021-09-08)

pipette 0.7.0 (2021-09-01)

pipette 0.6.3 (2021-08-05)

pipette 0.6.2 (2021-06-10)

pipette 0.6.1 (2021-06-09)

pipette 0.6.0 (2021-06-04)

pipette 0.5.18 (2021-05-18)

pipette 0.5.17 (2021-05-18)

pipette 0.5.16 (2021-05-18)

pipette 0.5.15 (2021-04-27)

pipette 0.5.14 (2021-03-17)

pipette 0.5.13 (2021-03-04)

pipette 0.5.12 (2021-02-22)

pipette 0.5.11 (2021-02-19)

pipette 0.5.10 (2021-02-17)

pipette 0.5.9 (2021-02-12)

pipete 0.5.8 (2021-02-12)

pipette 0.5.7 (2021-02-12)

pipette 0.5.6 (2021-02-11)

pipette 0.5.5 (2021-02-11)

pipette 0.5.4 (2021-02-09)

pipette 0.5.3 (2021-02-03)

pipette 0.5.2 (2021-01-15)

pipette 0.5.1 (2021-01-13)

pipette 0.5.0 (2021-01-13)

pipette 0.4.25 (2021-01-08)

pipette 0.4.24 (2021-01-06)

pipette 0.4.23 (2020-12-24)

pipette 0.4.22 (2020-12-11)

pipette 0.4.21 (2020-12-10)

pipette 0.4.20 (2020-12-09)

pipette 0.4.19 (2020-10-12)

pipette 0.4.18 (2020-10-09)

pipette 0.4.17 (2020-10-09)

pipette 0.4.16 (2020-10-06)

pipette 0.4.15 (2020-10-06)

pipette 0.4.14 (2020-09-14)

pipette 0.4.13 (2020-08-18)

pipette 0.4.12 (2020-08-13)

pipette 0.4.11 (2020-08-11)

pipette 0.4.10 (2020-08-03)

steinbaugh/brio
Pipette Biological Data In and Out of R

NEWS.md
In steinbaugh/brio: Pipette Biological Data In and Out of R