<<- in ternD(),
ternG(), and ternP(). Counter and tracker
variables are now accumulated via a local environment created with
new.env(parent = emptyenv()) rather than super-assignment,
which avoids unintended side-effects on parent frames (CRAN
policy).set.seed() inside ternG()
with withr::with_seed() so the global RNG state is never
modified (CRAN policy). Added withr to
Imports.\dontrun{} blocks replaced with
\donttest{} across all exported functions. Examples that
write files to disk use \donttest{} because the ‘Word’
export operations may exceed 5 seconds on slower machines; no missing
software or API keys are required.Description field of DESCRIPTION per CRAN
policy: ‘Word’, ‘Excel’, ‘tibble’, ‘officer’, ‘flextable’, ‘writexl’,
‘rstatix’.?TernTables) updated:
normality routing description corrected to four-gate algorithm;
.onLoad now explicitly initialises
TernTables.line_break_header option.val_format() now uses the Unicode plus-minus symbol (±)
for consistency with ternG() and ternD()
output.utils_naming.R: single-word variable names
(e.g. age, sex, race) now
correctly flow through the abbreviation map and capitalisation rules in
.apply_cleaning_rules().ternG() (e.g. Male/Female shown with
1.00 (ref.) and computed OR with 95% CI). Previously only
Y/N and 0/1 binary variables received ORs.factor_order = "mixed" option to
ternG() and ternD().utils_posthoc.R with center-based letter ordering (highest
center = “a”).getOption("TernTables.seed")).utils_preprocess.R to reduce false positives on common
research variable names.?TernTables).URL: and BugReports: fields to
DESCRIPTION.ternP(): preprocessing function for raw CSV/XLSX
data. Handles string NA conversion, whitespace trimming, empty column
and blank row removal, and case normalisation. Hard-stops on PHI column
name patterns and unnamed columns with data.write_cleaning_doc(): writes a ‘Word’ audit
document recording every transformation applied by
ternP().tern_colon_messy.csv to
inst/extdata/csv/ for use in ternP()
examples.line_break_header parameter refined: improved behaviour
for edge-case column widths in ternG(),
ternD(), and word_export().consider_normality = "ROBUST", now the default). Four-gate
algorithm applied per group: (1) n < 3 fail-safe to non-parametric;
(2) |skewness| > 2 to non-parametric; (3) all groups n ≥ 30 to
parametric via CLT; (4) Shapiro-Wilk p > 0.05 to parametric.
Implemented in both ternG() and ternD().ternB(): combines multiple TernTables result
tibbles into a single formatted ‘Word’ document.table_caption parameter to ternG(),
ternD(), and word_export(): places a bold
caption above the table in the ‘Word’ output.table_footnote parameter to ternG(),
ternD(), and word_export(): adds a merged
footer row below the table.line_break_header parameter to
ternG(), ternD(), and
word_export(): wraps long column headers onto two lines in
‘Word’ output.write_methods_doc() now generates a methods paragraph
tailored to ternB() multi-table output.ternG(): grouped comparison table for 2- and 3-level
group variables, with optional odds ratios (OR_col),
normality testing, and post-hoc test framework.ternD(): descriptive summary table with no group
comparisons.word_export(): exports any TernTables tibble to a
formatted ‘Word’ document.write_methods_doc(): generates a boilerplate
statistical methods paragraph.val_format() and val_p_format():
formatting utilities for publication-ready numeric and P value
display.
Need a high-speed mirror for your open-source project?
Contact our mirror admin team at info@clientvps.com.
This archive is provided as a free public service to the community.
Proudly supported by infrastructure from VPSPulse , RxServers , BuyNumber , UnitVPS , OffshoreName and secure payment technology by ArionPay.