NEWS.md

mark (development version)

• substantial changes to conditions

  • all conditions previously created with fuj::new_condition() have been rewritten with cnd::condition() [#259]
  • all of these condition now use different class names, generated through cnd::condition()
  • for more information, see https://jmbarbone.github.io/cnd
  • FIXME: ensure names: {rd_name}_{thing} -- and see about switching messages around: (message = function(type, params))

• the following arguments were deprecated and unused, and have now been removed completely:

  • list2df(show_NA)
  • vector2df(show_NA)
  • t_df(id)

• merge_list(sort) included to toggle name sorting in result

• make_sf() now uses substitute() to provide more information inside the function body when viewing #129

• round_to() added to round values to a set #142

• is_true(), is_false() now works as documented #262

• to_boolean() now works as documented #262

  • to_boolean.character(), and to_boolean.factor() have been improved; they will also now require exact matches; cleanup (e.g., trimming whitespace and lowercasing) are not longer performed
  • to_boolean.integer() added
  • to_boolean.numeric(), to_boolean.integer() will now return NA

• md5(bytes) added to use tools::md5sum(bytpes) for R > 4.2.0 #258

• md5() now uses little-endian serialization (i.e., serialize(xdr = FALSE)) for more consistent results across platforms and faster speed; which may cause hashes created prior to change

• {mark}'s title has been updated

• write_file_md5(method = "feather"), write_file_md5(method = "parquet") now use {feather} and {nanoparquet}, respectively, rather than {arrow} #245

• read_clipboard("md") no longer needs {readMdTable}

• read_clibpboard() will always return a tibble if {tibble} is available (this can be turned off if options(mark.tibble = FALSE))

mark 0.8.3

• file_copy_md5() now produces more messages #239
• file_copy_md5() saves md5 sum checks as an attribute
• removes tests for struct() #252

mark 0.8.2

• read_clipboard() and write_clipboard() now use {clipr} to work on non-Windows platforms #125
• read_clipboard() now works with more methods for reading data.frames
• read_clipboard() now defaults to a tibble return when {tibble} is available
• timezone testing updated #247
• actions updated

mark 0.8.1

• write_file_md5() now supports "feather" and "parquet" methods as wrappers for {arrow} #234
• md5() added to provide MD5 check sums for objects #233
• unique_rows() added to subset on (non-)duplicated rows in a data.frame #87
• within() added as an alternative to between_more() #120
• test updated for upcoming R release #240

mark 0.8.0

breaking changes

• echo() is now removed; use echo::echo() instead #214
• includes tryn() for running an expression a maximum number of times before failure #80

fixes

• unlist0() no longer fails when input list is not named #220

improvements

• match_param() has been improved
  • can now return multiple matches #191, and can return partial matches
  • error message readability improved for matchParamMatchError#194
  • choices can now be a list of formula elements, preserving the return value

new features

• file_copy_md5() added as a wrapper for fs::file_copy() but provides MD5 checks through tools::md5sum() to avoid overwriting files that had no content changes #207
• write_file_md5() added as a general writing function and utilizes file_copy_md5() for MD5 checks (including some compression options) #207, #224

mark 0.7.0

• merge_list() added for combining lists #200
• glob() added for basic wildcard globbing on character vectors #167
• adds greater use of {fs} over base file functions #160
• improvements in todos() and fixmes()
  • File extension can now be set #170, which by default includes qmd (#163) and py files
  • new parameter ignore to ignore any files
  • file paths and line numbers can now be clicked within RStudio #171
• adds more use of rlang::list2() for internally #199
• GitHub action included to check version updates with pull requests #211
• %::% and %:::% now exported from {fuj}

mark 0.6.1

• updates Timezone references for upcoming R release #203

mark 0.6.0

New features

• new functions for detecting blank values in a vector or data.frame. Blank values are those which do not contain any text (controls for NA) or are entirely white space.
  • is_blank() for detecting blank values in a vector
  • is_blank_cols() for detecting blank columns
  • select_blank_cols() for selecting blank columns
  • remove_blank_cols() for removing blank columns
• match_param() now accepts a named listed for alias matching #104
• echo() evaluates expressions and logs outputs #164
• {fuj} is now imported
  • multiple functions now re-exported from {fuj} (see ?mark::reexports)
  • set_names0() is deprecated in favor of set_names()
  • error messages are created with fuj::new_condition();
  • test for errors and warnings enhanced with class checks

Fixes and updates

• date_from_partial() works again #155 after fixing an issue with an internal utility is_valid_date_string() that wasn't recognizing %Y-%m-%d (and potentially others)
• lintr GitHub action updated #173; this includes plenty of internal improvements and code cleanup
• package description in help files corrected #165
• GitHub pages updated with latest {pkgdown} action #175
• Update to GitHub R-CMD-check action #178
• switch_in_case() handles NAs better #183
• internal switch tests updated for {waldo} development #182 thanks, @hadley
• methods for write_clipboard() are now displayed in documentation #186

mark 0.5.3

• CRAN fix for new release #151

mark 0.5.2

New features

• normalize() added to normalize values in vectors, matrices, and data.frames by specified ranges and boundaries #143
• get_labels() and other label related functions now get exact matches for "label" attributes #141
• recode_only(), and recode_by() now accept a named list() for by #96]
• switch_in_case() now handles functions in the right hand statements
• update_version() now correctly checks result of embedded utils::menu() call for updating the version #123
• require_namespace() now accepts multiple namespaces #121
• unique.fact() S3 method #86
• recode_only() and recode_by() can accept a single value for val #73
• fact_reverse() for reversing fact levels #78
• as.Date.fact() added #108
• as.character.fact() added
• [.fact added
• read_bib() better handles fields where = is present in the text #117

Fixes

• fact.haven_labelled() works properly and retains the label attribute #136
• drop_levels() is exported #105
• recode_by() and recode_only() handle factors better #81
• Functions that made use of shell.exec() now try to determine the appropriate method of opening a file base on OS. #126
• Internal functions for potentially coercing factor levels into dates no longer try to check for "%Z" in the date format #147

Breaking changes

• reverse() has been removed (use flip() instead)
• assign_label() has been removed (use assign_labels() instead)
• percentile_rank(times) is deprecated in favor of percent_rank(weights)

Non visible changes

• print.fact() rewritten as a slightly modified print.factor() #109
• percentile_rank() improvements #131

mark 0.5.1

• Fix for CRAN check #128

mark 0.5.0

The package website https://jmbarbone.github.io/mark/ is now available! More references and vignettes will make their way here in future releases.

Fixes

• detail(NA) (or when x has only NA values) no longer throws a warning and returns NA for min_c, max_c #59
• print.noted() now passes ... to next methods #67
• corrects deprecation warning in assign_label() #74
  • assign_label() will be removed in 0.4.2
• set_not_available() now seems to work correctly -- it probably hasn't actually be working most of the time
• percentile_rank() is now more correct when x is a decimal by checking for unique values first #92
• counts.data.frame() now handle factor columns better

New features

• Functions in ?handlers, all allow for additional params passed through ... #34
• adds row_bind() to bind a list of data.frames() #46
• adds drop_levels() with factor and data.frame methods; functions similarly to base::droplevels() but is a little faster #54
• todos() and fixmes() gain a new param force
  • When TRUE, forces searches for .R files when the given directory does not contain an .Rproj file
  • When FALSE, prevents long start ups when these functions are called in a .Rprofile file and R is not launches in a project directory
  • This be toggled with a new options mark.todos.force
• adds set_note(), a wrapper for note<-() #77
• adds fact2char() to compliment char2fact() #75
• print.pseudo_id() now truncates long uniques to a single line #70
• match_param(NULL, null = TRUE) allows param to safely return NULL #89
• fact_na() is added to use fact vectors with NA levels that work with is.na() #69 and other NA handling improvements
• adds a new print._mark_bib_df() method to supporting printing lists
• adds new methods for facts: as.integer.fact(), as.double.fact(), remove_na.fact()

Breaking changes

• fact.numeric() now treats NaN the same as NA, no extra level/unique value is retained
• read_clipboard() now returns NA when the clipboard is empty, rather than "" (improvements with internal type conversions)
• improvements to NA handling as well

Other, non-visible

• github actions updated
• internal type conversion now heavily relies on utils::type.convert() with some additional functionality for logical (e.g., character string using "true" and "false") and for guessing dates in a YYYY-MM-DD format
• general clean up and formatting

mark 0.4.1

• details(factor) no longer adds fact class to factors #50
• details() gains new argument factor_n to control threshold for making character vectors into factors
• detail.data.frame() now works with single column data.frames #48
• paste_combine() no longer duplicated the second vector of ... when length(...) > 2 #52

mark 0.4.0

New features

• adds percentile_rank() to calculate percentile ranks with a vector
• adds insert() to insert multiple values into a vector
• pseudo_id() gains argument na_last to change positioning of NA values
• is_true() and is_false() are now exported as generics with methods for default and logical
• adds omit_na() for tracking positions of NA and non-NA values
• quick_df(NULL) now returns an empty data.frame
• quick_dfl() exported as a wrapper for quick_df(list(...))

Fixes

• squash_vec() now works correctly when values are not ordered #43
• as_ordered() no longer duplicates ordered class #44
• counts.data.frame() and props.data.frame() correctly make column names unique
• internal try_numeric() correctly handles NAs
• flip.matrix(, keep_rownames = FALSE) now works correctly
• any_match() now works as expected
• lines_of_r_code() now works correctly reading a single file
• import(, overwrite = TRUE) now works
• ls_function(), ls_object(), ls_dataframe(), and ls_all() have improvements for environmental searching
• assign_labels.data.frame(.missing = "warning") correctly removes missing columns
• remove_na.factor() no long drops additional classes other than ordered and factor

Others

• documentation of struct() overwriting attributes improved and examples
• adds more unit tests

mark 0.3.0

Fixes

• fact.haven_labelled() now returns an object with class fact #39; performance enhancements
• set_names0(NULL) no longer causes an error and returns NULL #40
• diff_time() correctly handles time zones when x is Date and y is POSIXt #41

Changes

• updates file path finding functions (e.g., list_files()) to try to not search every file depending on desired searches (e.g., by full file paths or just base names)
• as_ordered() handles factors better; S3 methods removed: as_ordered.ordered(), as_ordered.factor()
• remove_na() has better performance when x has no NA values
• counts.data.frame() and props.data.frame() retain attributes of selected columns
• todos() and fixmes() will not search for .R or .Rmd files if the path is not changed from "" and no .Rproj is found within the directory

New features

• adds unlist0() to retain original names of lists
• adds %names% for a fun way to set names
• adds file_open() as alias for open_file()
• adds detail() to return a data.frame of details for a vector of columns of a data.frame
• adds squash_vec() to combined the names of a vector with repeated values
• adds make_sf() as a simple wrapper for package specific system.file()
• add_file_timestampe() gains a new parameter sep to separate the file name (sans ext) and the time stamp
• assign_labels.data.frame() gains new argument .ls to explicitly set a list (or data.frame) of columns
• props() and family gain argument na.rm to not count NA values for proportions

mark 0.2.0

Changes

• package_available() now visibly returns TRUE/FALSE
• remove_na() now has methods for lists and factors
• environments() now has it's own print.mark_environments() method rather than calling cat() within the function itself
• array_extract()'s first argument is changed from arr to .arr
• diff_time() now defaults to using UTC (Related to #32)
• print.note() method has been updated (Related to: #33):
  • to print x normally, without the note class when just the note has to be seen
  • an internal function now handles the note formatted for class noted
• changes to fact()
  • fact() now returns a vector with a fact element
  • fact.character() correctly labels NAs #24
  • fact.factor() not longer simply returns x but rather updates the levels and integer values to confirm with other fact() methods. fact.factor() will retain all levels of the original value, reorder the levels, and append NA if necessary
  • fact.fact() added to return a correctly formatted fact()
  • fact.logical() now orders levels as TRUE then FALSE, and NA if present
  • fact.Date() and fact.POSIXt() added, which simply call fact.numeric()
  • print.fact() method added to print a fact vector as a factor
  • as_ordered.factor() and as_ordered.ordered() now call fact() to check levels

Fixes

• functions that check if an argument is a vector no long use is.vector() directly; arguments passed with attributes that when removed fulfill is.vector() are accepted
• todos() and fixmes() now correctly show tags for .Rmd files
• correction to error message in limit()
• adds missing sort argument to props()
• pseudo_id.factor() no longer returns NA_integer when a value is NA or a level is NA and correctly resets the order of the levels from the factor to their order of appearance
• flip.data.frame() no longer coerces single column data.frames #36

New features

• fact.pseudo_id() and pseudo_id.pseudo_id() methods added
• adds as_ordered() to quickly create ordered factors using fact()
• adds char2fact() to convert character vectors (or columns in a data.frame) to factors based on the number unique values
• adds tableNA() to make a table from NA values
• round_by() gains an additional argument include0 which if FALSE will replace 0 values with by
• assign_labels.data.frame() gains an additional argument .missing to set how to control for missing labels: you can now use a warning for a missing label (instead of an error) or silently ignore any missing labels
• sort_names() gains a new argument numeric to try to sort names of x by their numeric value #26
• adds struct(), a simplified version of struct()
• adds more fact() methods
  • fact.integer() #30
  • fact.haven_labelled() #31
• todos() and fixmes() gain an additional argument path to specify a directory or file to search within #25
• print.pseudo_id() added for a cleaner print
• between_more() accepts vectors for left and right params

Non visible changes

• code coverage added
• additional tests added

mark 0.1.4

• no visible user changes
• removes temporarily created files #22

mark 0.1.3

New name! The previous name jordan was conflicting with recent package on CRAN.

Changes

• corrects use of ... in todos() #8
  • grep() also now evaluated with cleaned todo text
  • searches for todos() in Rmd files, too
  • correctly removes additional # and spaces in lines (e.g., # # TODO text -> text)
• updates for counts()
  • corrects NA counting in counts(); NA counts are now appended at the end whether or not sort is called
  • other optimization for counts()
  • core functions previously on base::rle() now use a combination of pseudo_id() and base::tabulate()
  • corrects counts for factor data when higher levels are not present #16
• update to multi_grepl() internal functions to prevent conflicts with R 4.1.0
• corrects error message in vector2df() when passed a list

New features

• adds/exports fact() and pseudo_id()
• adds fixmes() #13
• adds names_switch() to switch names and values
• vector2df() can now output a 1 column data.frame if name = NULL
• adds an invert parameter to complete_cases() to filter for incomplete cases
• adds are_identical() for comparing 2 or more vectors as identical(), element-wise
• adds add_file_timestamp()
• diff_time() and related functions will try to convert y to a Date object if x is passed as date (e.g., diff_time_days(Sys.Date(), "2021-06-03") will not show decimals)

jordan 0.1.2

Changes

• removes checks for stringsAsFactors option in quick_df()
• improves functionality of note
  • removes jordan.note.fun option for printing -- this was too complicated and doesn't seem to work too well without too many adjustments
  • note<- now appends the class noted to the object so that a print.noted method is dispatched so the note will be printed when called
  • print.note note defaults to a colorful message called with cat()
  • Startup related functions moved to jordanExtra; these were a bit wild, dynamic, and not well tested. The .Rprofile template also exists in the separate package.

Improvements

• match_param() now reports the value passed to param on failure
• improvements to handlers
  • adds has_message() and get_message()
  • internal rework of catch() for catching errors, messages, and warnings
  • has_*() now returns the result in the result attribute
  • print.has_catch() will hide attributes in print
• improvements/updates to todos()
  • allows text filtering by passing arguments to grep()
  • adds new print method for todos_df for viewing tasks
  • the result of todos() still has class data.frame but will now also have class todos_df
  • the print.todos_df() method should be make more sense for task management
• str_slice_by_word() no longer has a leading " " for each element after the first (this was not the intention of the split)
• is_file() and is_dir() now returns and error when passed NULL or a vector of length 0
• switch_params() now accepts a vector for x
  • ... examples updated
• adds param to change default column name from counts.data.frame() and props.data.frame()
• print() method called from todos() has a new format to group together multiple items found in a single file

New features

• adds print_c() to print a vector into a c() "string"
• adds diff_time functions
  • functions include diff_time() along with shortcuts for specific methods: diff_time_secs(), diff_time_days(), diff_time_years(), and others (see ?jordan::diff_time)
  • these are much like the base::difftime() but...
    • have more methods/units for computing differences, which may need to be the case for when a unit of time has to be standardized (e.g., 1 year needs to be 365 days, or 1 month needs to be 30 days)
    • are class diff_time and numeric
    • have a slightly different print method (will note appropriately how units of time are measures)
    • can account for timezone differences (must be set with a the tza and tzb parameters)
• adds sort_by()
• adds NA assignments:
  • NA_at() for position assignments
  • NA_if() for logical/conditional assignments
  • NA_in() for inclusive matching assignments
  • NA_out() for exclusive matching assignments
• adds functions for sourcing scripts into environments for later use (rscript(), save_source(), and source_to_env())
• adds switch_case() to return a values based on a left hand statement returning TRUE and switch_in_case() for evaluating x %in% left hand side
  • these functions are much like dplyr::case_when() but for specific cases
  • adds function checkOptions()
• adds recode_by() and recode_only() for a simple implementations of recoding elements in a vector

jordan 0.1.1

New features

• adds fizzbuzz()
• adds data.frame functions
  • adds quick_df() to turn a list into a data.frame (used internally, too)
  • adds complete_cases() to select rows without NA values
  • removes show_NA parameter from vector2df() and list2df()
    • for vectors this will now produce an NA value for the first column
    • for lists make.unique() is utilized for empty name named to retain the position of the list element
• adds listing wrapper:
  • ls_object() to list all is.object()s
  • ls_dataframe() to list all is.data.frame()s
  • ls_function() to list all is.function()s
• adds counts() and props() for counting unique elements in vectors and data.frames

Changes

Some exported functions and names have been changed to prevent conflicts with other popular packages

• %||% is no longer exported; it is exported in rlang (and reexported in purrr) and is a relatively simply function anyway
• collapse() is now collapse0() to avoid conflicts with glue; although glue::collapse() is meant to be deprecated, collapse0() is mostly a wrapper for paste0(), so this may be a better name
• set_names() is now set_names0() to avoid conflicts with rlang (reexported from purrr) and magrittr

Fixes/updates

• do_paste_combine() (used inside paste_combine()) simplified to remove use of outer()
• improves version bumping/updating
  • added get_version() to retrieve the current package version (assuming you're in the directory)
  • utils::menu() is called to confirm that version should be updated
  • can update by either adding a number to the version (bump_version() or by date bump_date_version())
• implements an improved non-exported string_extract() function inside str_extract_date() and str_extract_datettime()

jordan 0.1.0

Major cleanup for documenting, reviewing, removing, relocating, and testing functions.

• Added a NEWS.md file to track changes to the package
• Files renamed and reorganized
• Various tests included
• Added aliases to muffle messages and warnings (muffle() and wuffle())
• Added limit()
• match_param() now returns correct errors messages for calls
• write_clipboard() formats vectors as characters
• read_clipboard() tries to correctly format vectors and data.frames
• removed str_close_enough() because this didn't make much sense anyway

New functions

• %colons%: A substitute for :: and :::
• %||%: Virtually identical to rlang's version
• .CharacterIndex(): Quick visual of indexes in character strings
• bump_date_version: Updated DESCRIPTION versions that are saved as dates
• chr_split(): Essentially an alias for strsplit(., "")[[1]]
• Error and warning handlers: get_error(), get_warning(), has_error(), has_warning(), muffle() and wuffle() (aliases for suppressMessages() and suppressWarnings())
• File/directory utilities: is_dir(), is_file(), list_dirs()
• is_na_cols(): previously not exported, inside select_na_cols() and remove_na_cols()
• limit(): Limits a numeric vector by an upper and lower bound
• match_param(): Alternative to match.arg() without partial matching and more detailed error message
• Functions for calls: outer_call(), outer_fun(), within_call(), within_fun()
• quiet_stop(): calls stop() without throwing an error
• Functions for names: set_names(), remove_names()
• require_namespace(): Mostly a wrapper for require() with a more detailed error message
• vap_*(): Wrappers for vapply(), must like purrr's map_*

Moved to jordanExtra

Some miscellaneous, less controlled functions have been moved to jordanExtra.

• Functions for Rust set_rust_engine(), engine_rust()
• Functions for openxlsx: add_data_sheet(), add_image_sheet()
• Functions for pROC: pROC_optimal_threshold(), pROC_quick_plot()
• Effect sizes: cohen2odds(), cohend2r(), odds_ratio(), odds2d(), odds2r(), r2cohend()
• Statistical functions: fishers_method(), iqrs(), p_round(), p_value_sig(), percentile_rank(), proportion(), sd_pooled(), sterr(), tukey_coef(), z_score()
• Others: add_euclidean(), add_malahanobis(), %=+, filter_combine(), reverse_log_trans(),

0.0.9000

• Initial commits of functions
• Quite a mess with no versions