footnote_plan() receives a new argument,
order, allowing users to specify the order of footnotes
(#605, @alanahjonas95).row_grp_plan()/row_grp_structure() to remove
trailing post-space rows (#630, @alanahjonas95).extract_data() function to enable data to be
easily extracted from a tfrmt into a data frame (#628, @alanahjonas95).shuffle_card() now automatically strips inherited ARD
attributes from objects of class bind_ard to avoid
unreliable metadata (#650).shuffle_card() now prioritises the user-supplied
by argument over inherited ARD attributes, giving users
control to override attributes and updating the mismatch message to
reflect this change (#650).tfrmt_sigdig so it correctly passes the
‘missing’ argument to the body_plan (#621, @alanahjonas95).page_plan where the max_rows argument was
returning an error if a value in the group variable was an empty string
(#539, @alanahjonas95).row_grp_plan where identical labels repeated
across different groups were incorrectly forced to the top of the table
instead of maintaining their correct order (#634, @alanahjonas95).as.character.span_structure where the regex
was incorrectly capturing content wrapped in parentheses as an R
function even though it did not have a valid R function identifier
immediately before the (. i.e., “n (%)” incorrectly
captured as a function (#643, @LiamHobby).page_plan() receives a new argument,
transform, allowing users to modify the page label (#496,
@dragosmg).Patch release for latest {gt} release and upcoming {purrr} and {stringr} releases:
rowname_col = NULL in gt() within
print_to_gt() if no row label exists.seq_along(x) in lieu of 1:length(x)
prior to str_replace() as it no longer accepts NA
patterns.map_chr() as it no longer coerces to character.shuffle_card(), which prepares a
card ARD object for analysis.prep_...() functions to reduce the manual
processing required for a card ARD object once it has been
shuffled. (#509, @dragosmg)
prep_combine_vars(): useful for combining sparsely
populated columns into a single one.prep_big_n(): useful when the tfrmt contains a bigN
structure. It recodes the "n" stat_name into
"bigN" for the desired variables.prep_label(): combines stat_label and
variable_level for categorical (or categorical-like)
variables.prep_hierarchical_fill(): does a pair-wise replacement
of missing values (NAs) in one column based on another
column. It can replace NA either with a predetermined
value, with the value of the other column in the pair, or with a {glue}
expression.*_structure functions did not correctly
parse strings containing quotes in some cases. (#466)col_style_plan in the creation
of mock data.frmt_when dropped
quotes from stringsdplyr::summarise()body_plan may error if the
grouping variable is a factorfrmt_combine couldn’t process
variable names surrounded by backticksrow_grp_plan post space did not
respect col_style_plan widths by adding new
fill argument to element_block. The
fill argument controls whether post space values should be
recycled for the cell’s data width. For example, a cell width of 3 will
be respected by the post space with the following syntax:
element_block(post_space = "---", fill = FALSE).border argument in
element_block.row_grp_plan splits on all grouping
variables, even if not mentioned. Instead, the logic has been updated to
split on those explicitly mentioned, similar to
page_plan.unicode_ws argument added to
print_to_gt() and print_mock_gt(). This
defaults to TRUE but should be set to FALSE
for RTF outputs (until {gt} bug is resolved).make_mock_data could result in
duplicate rowscol_style_plan). Alignment options now fall into two types:
character (type = “char”) and positional (type = “pos”). Positional
alignment is new and allows for aligning across multiple positions.page_plan for splitting tables across multiple
pagescol_planrow_grp_plan that
stores row_grp_structures has been changed from “struct_ls”
to “struct_list” to be consistent with other objects. This may impact
compatibility with JSON files created using prior versions of
{tfrmt}.frmt_combine no longer throws error if group variable
is named “var”row_grp_plan with post-space no longer throws error if
character column contains NA valuesfrmt()NEWS.md file to track changes to the
package.