• clear some notes and warnings in preparation for CRAN submission (#265)
  • bug fix to read escaped enclosure characters in free text items, e.g. \" (#220, #261)
    • read_export_options() now also reads enclosure character from header
    • .print.secuTrialoptions() now also prints enclosure character. Quote and Tabulator are now printed as '\"' and '\t' rather than '"' and ' '.
    • read_export_table() now uses readr::read_delim() with options escape_backslash = TRUE and quote = export_options$quote
  • warning for ISO-8859 encoding moved from read_export_table() to read_secuTrial_raw() (one warning per export file instead of several warnings for each table)
  • read_export_options() now reads only the first 10 lines to find the encoding. Reading all lines would fail if the export options were ISO-8859 encoded and contained ‘problematic’ characaters, e.g. in center names.
  • Removed encoding option in plot_recruitment(): Not needed here, as centres are read with correct encoding in read_secuTrial_raw() and stored in the secuTrialR object.
  • write_secuTrial() no longer allows format = "sas" due to deprecation of write_sas in haven (version 2.5.2).
  • read_secuTrial_raw now produces a more explicit warning when ISO-8859 encoding is used. Recent changes in R seem to have reduced compatibility with this encoding.
  • fix CRAN notes - remove codecov badge, check class via inherits
  • warning regarding overwritten pat_id variable
  • adapt as.data.frame.secuTrialData() to allow named character vectors as option data.frames to specify custom names for data.frames (#250)
  • recommendation to export with short names (#238)
  • addition of arguments in read_export_table etc allowing passing other options to read.table (#240)
  • addition of dictionary_secuTrial function for easy access a data dictionary (#241)
  • formtablename added to the output of visit_structure for better reference between the table names and labels (#242)
  • addition of export_date and export_datetime for easier access to the export date (and time) (#246)
  • adapt read_export_options() to work with exports from new secuTrial version (6.1.2.5 - 6.3.2.5) (#245)
  • transferred the Maintainer tag to Alan Haynes
  • reverted factorize_secuTrial() back to 1.0.3 version due to problems with lookup table factorization (#224)
  • added citation (doi: 10.21105/joss.02816)
  • moved tcltk and igraph dependency to suggested (#223)
  • improved import speed, specifically through changes in dates_secutrial() and factorize_secutrial() (#204)
  • adjustments to handle review feedback from CRAN (#190)
  • restructuring in preparation for a release on CRAN
  • added suggestion to NOT export form data of hidden fields (#177)
  • extended failure comment in read_secuTrial() to indicate that the problem could be a rectangular export file (#168)
  • added “Form data of hidden fields” export option information to export_options (#171)
  • added return_hidden_items() function (#172)
  • bug fix: presence of the audit trail was incorrectly identified due to a comment in the source file of the export options (see #155, comments from @suvi-subra and @OliviaEbnerIAS)
  • Added sorting option to visit_structure. (#152)
  • added up-to-date vignette (#99)
  • path in print.secuTrialdata is now wrapped at 80 characters
  • secutrialoptions class is now secuTrialoptions.
  • add appveyor testing, pkgdown site
  • fix possible bug on windows due to regex in .prep_line_items (used in plot_recruitment) (#147)
  • Changed license for the package from GPL-2 to MIT.
  • The general nomenclature for a study subject will from now on be participant (pat). All variations of this (e.g. case, patient) have been adjusted in the code and the documentation.
  • Removed generic plot() function for secuTrialdata objects. (#139)
  • dates_secuTrial() incomplete date warnings are now concatenated and returned as one warning per form instead of many. (#124)
  • Fixed issue #121 on GitHub. factorize_secuTrial() can now handle exports which have the reset option enabled in radio buttons.
  • check_export_options() function was added. It informs on deviations from suggested export options. (closes #17)
  • Removed tracking of obsolete export options (partial_date_string, partial_date_handling, unknown_date_string).
  • Added format_info (e.g. “CSV format for MS Excel”) to export_options.
  • Fixed issue #116 on GitHub.
  • subset_secuTrial() function was added. It allows subsetting of secuTrialdata based on patient ID and/or study centre name.
  • get_participants() function was added. It allows easy extractions of participant info from a secuTrialdata object.
  • return_random_cases() now returns a list. The first element are the cases and the second element is the output of RNGkind().
  • New function diff_secuTrial() added to allow light weight comparison of the setup of two secuTrial exports.
  • Metadata variables are now also transformed to date and datetime formats, whenever appropriate.
  • factorize_secuTrial() now no longer triggers an unexpected warning when the name of a secuTrial lookuptable is equal to the name of the variable it is being used in. (PR #108)
  • return_random_cases() has been added to the package. It allows to sample a random subset of cases from a secuTrial export in a reproducible fashion.
  • read_secuTrial_raw() and read_secuTrial() no longer fail due to missing Add-ID, centre information or project setup in export data. Instead, adding of pat_id (no Add-ID), centre (no centre information) and visit_name (no project setup) to the data tables is now omitted if the relevant data for the operation is not available.
  • dates_secuTrial() now warns if not all dates were parsed (expected if there are incomplete dates).
  • factorize_secuTrial() now warns if there are issues with the factorization (not expected to trigger).
  • New function build_secuTrial_url() has been added. It allows users to easily compose URLs to specific secuTrial forms.
  • As of version 0.4.17, changes will be recorded in the NEWS file.