Tplyr 1.4.1

A single bug fix release.

Bug fixes

• An unnamed empty in f_str() was ignored when every value on a row was missing (#215). In a descriptive statistics layer, a row where all summarized values are NA came out blank instead of filling the format string.

  group_desc(AVAL, by = PARAMCD) %>%
    set_format_strings(
      'n'        = f_str('xxx', n, empty = "0"),
      'mean, sd' = f_str('xxx.xx, xxx.xx', mean, sd, empty = "NA")
    )

  row	before	after
  n	(blank)	0
  mean, sd	(blank)	NA, NA

  A regression introduced in 1.3.2 when descriptive string construction was vectorized. The replacement dropped all-missing rows from formatting entirely, but only populated them when empty was named .overall — so an unnamed empty fell through both branches.

  Behaviour for the other two empty modes is unchanged: the default c(.overall = '') still blanks the cell, and a named c(.overall = "NONE") still replaces the whole result.

---

Full Changelog: v1.4.0...v1.4.1

Tplyr 1.4.0

New Features

• Overall maximum precision for auto-precision formats (#219). set_format_strings() gains max_int and max_dec arguments, backed by a new tplyr.max_precision option. The existing cap argument limits the length allotted to the a of an auto-precision format, but the + modifier is applied on top of it, so a+2 could still grow past the cap. The new maximums are applied after auto-precision and the + modifier combine:

  group_desc(AVAL, by = PARAMCD) %>%
    set_format_strings(
      'Mean (SD)' = f_str('a.a+1 (a.a+2)', mean, sd),
      max_dec = 4
    )

  They apply only to format groups that use auto-precision — an explicitly specified format like xx.xxx is left alone. max_int follows the existing integer semantics: it limits the space allotted, but integers never truncate. Defaults reproduce previous output exactly.

Bug fixes

• set_shift_layer_formats() table level defaults silently ignored (#216). Shift layers that didn't set their own format strings fell through to the option default, rendering counts unformatted and dropping pct entirely.
• Incorrect add_missing_subjects_row() counts on nested count layers (#217). set_distinct_by() was not applied to the inner layer, so counts reflected distinct outer groups rather than distinct subjects.

Both were regressions introduced by the 1.3.2 internal refactor, with the same root cause: an environment read that no longer inherited from the parent, failing silently rather than erroring.

---

Full Changelog: v1.3.2...v1.4.0

1.3.2

Tplyr 1.3.2

Internal Changes

• Refactored internal functions to eliminate evalq() usage and adopt the Extract-Process-Bind pattern, improving code clarity and maintainability without affecting user-facing functionality
• Performance optimizations for nested count layers using vectorized operations (~2.4x speedup)
• Optimized count formatting and prepare_format_metadata() internals
• Removed unused un-exported functions to reduce code base size
• Updated GitHub Actions workflows

Bug fixes

• Resolve #202 Fix VARN sorting for nested count layers
• Resolve #193 Fix str_extract_num() to properly identify negative numbers
• Resolve #186 Update vignettes to reset options back to defaults to avoid CRAN inconsistencies
• Resolve #184 Don't remove units attribute of difftime when cleaning data on pre-process
• Resolve #182 Fix error when adding multiple risk difference comparisons at once with NAs in two-way matrix
• Resolve #183 Insert row breaks between layers without removing duplicated

Tplyr 1.2.1

Release 1.2.1 is a quick patch follow up to 1.2.0

Tplyr 1.2.1

• Resolve #178 to add metadata handling for missing subjects, and add the add_anti_join() function

Tplyr 1.2.0

• Resolve #62 Add data vignette data into the package (thanks for the suggestion @thebioengineer)
• Resolve #74 Add an example of piping in set_pop_data
• Resolve #83 Add the add_missing_subjects() function
• Resolve #84 Add set_limit_data_by() function
• Resolve #111, #148 Allow ellipsis argument unpacking outside of add_layer()
• Resolve #129 Add collapse_row_labels() function
• Resolve #134 Clarify how population data works to trigger denominators
• Resolve #75, #146, #166 Fix nested count layer handling where one inner layer value exists in multiple outer layer groups
• Resolve #21, #152 Fix handling of Inf, -Inf in desc layer for min and max
• Resolve #154 Fix namespace scoping for execution of Tplyr tables within non-global environments
• Resolve #155 Dead code clean-up
• Resolve #170 Add replace_leading_whitespace() post-processing function
• Resolve #173 Fix nested count layer sort variable behavior when using by variables

Thank you to @asbates @ShiyuC @alexhomel @om9391 @elimillera

Tplyr 1.1.0

What's Changed

• Fix custom metadata vignette #113 and increment dev version by @mstackhouse in #114
• Gh issue 115 by @mstackhouse in #116
• Add Scrollbar for Articles by @mattroumaya in #119
• Issue #117 - Add Parenthesis Hugging by @mstackhouse in #118
• Update snapshots for current package versions by @mstackhouse in #121
• 1.1.0 Release by @mstackhouse in #120
• Update workflow versions by @elimillera in #122

New Contributors

• @mattroumaya made their first contribution in #119

Full Changelog: v1.0.2...v1.1.0

Tplyr 1.0.2

What's Changed

• closes #99 closes #101 doc: fix typo in multiple files by @sadchla-codes in #100
• Closes #103 Closes#104 tplyr reference update@devel by @sadchla-codes in #105
• Update logic for determining denom_where by @elimillera in #107
• Introduce bug fix for #106 to prepare for CRAN release by @mstackhouse in #108
• Update pkgdown.yaml by @elimillera in #109
• Update pkgdown workflow by @mstackhouse in #110

New Contributors

• @sadchla-codes made their first contribution in #100

Full Changelog: v1.0.1...v1.0.2

Tplyr 1.0.1

• Fixes an issue with revdeps

What's Changed

• Update pkgdown action to use main by @elimillera in #94
• Gh issue 95 by @mstackhouse in #96
• Devel by @elimillera in #97

Full Changelog: v1.0.0...v1.0.1

Tplyr 1.0.0

Tplyr 1.0.0

• New features
  • Introduction of traceability metadata framework #32
  • Framework for creating re-usable layer templates #66
  • Native pipe compatibility #33
  • Automatically load magrittr pipe #22
  • Refactor of count layer programming #28
  • Allow external precision data for desc layers #27
  • Allow denominators within count layers as formattable values #11
  • Descriptive statistics layers allow stats as columns #37
  • New experimental function set_numeric_threshold()
  • Apply f_str's outside of a Tplyr table with new function apply_formats() #57
  • New post processing function helper str_indent_wrap() for hyphen-enabled string wrapping #59
• Bug fixes
  • Fix errors in partially provided precision caps #20

v0.4.4

Tplyr 0.4.4 Release