Package 'TernTables'

A data frame or tibble.

Optional character vector of variable names to assess. If NULL (default), all assessable numeric variables are included (excluding exclude_vars and group_var).

Optional character vector of variable names to exclude.

Optional name of the grouping variable (as used in ternG()). When provided, the normality assessment is performed across groups simultaneously — exactly as ternG() does. When NULL, each variable is assessed as a single vector (matching ternD()).

Normality assessment mode — must match what was (or will be) passed to ternG() / ternD() to guarantee identical routing. "ROBUST" (default), TRUE, or FALSE.

A ternP_result object returned by ternP.

Currently unused; included for S3-method compatibility.

A list of tibbles created by ternD() or ternG(). Must be constructed with list(), not c() (e.g. list(T1, T2, T3)). Each tibble must have been produced in the current R session; the metadata is stored in memory, not in the tibble columns.

Output file path ending in .docx.

Logical; if TRUE (default), inserts a page break between each consecutive table.

Logical; if TRUE, writes a single methods section Word document that covers all tables in the list. Statistical test details are pooled across all tables. Default is FALSE.

Output file path for the methods document. Defaults to "TernTables_methods.docx" in the working directory.

Logical; if TRUE (default), automatically opens each written .docx in the system default application after saving. Set to FALSE to suppress.

Logical; if TRUE (default), appends a citation line at the bottom of each table footnote block and methods document: package version, authors, and links to the GitHub repository and web interface. Set to FALSE to suppress.

Character; font family for all Word output. Any font name accepted by the rendering system is valid. Can also be set via options(TernTables.font_family = "Garamond"). Default "Arial".

Tibble with variables.

Character vector of variables to summarize. Defaults to all except exclude_vars.

Character vector to exclude from the summary.

Character vector of variables to treat as ordinal (i.e., use median [IQR]) regardless of the consider_normality setting. This parameter takes priority over normality testing when consider_normality = "ROBUST" or TRUE.

Character vector of variable names to treat as normally distributed, bypassing all normality assessment. Listed variables are summarized as mean $\pm$ SD regardless of the consider_normality setting. Takes priority over consider_normality but not over force_ordinal (if a variable appears in both, force_ordinal wins). Default is NULL.

Character vector of variables to force treatment as continuous (mean $\pm$ SD), bypassing the automatic binary 0/1 detection that would otherwise convert them to categorical Y/N. Useful when a numeric variable with only two unique values (e.g. 0/1 dose levels) should be analysed as a continuous measurement rather than a dichotomous category. Default is NULL.

Optional Excel filename to export the table.

Optional Word filename to export the table.

Character or logical; controls routing of continuous variables to mean $\pm$ SD vs median [IQR]. "ROBUST" (default) applies a four-gate decision: (1) n < 3 $\rightarrow$ non-parametric (conservative fail-safe); (2) absolute skewness > 2 or excess kurtosis > 7 $\rightarrow$ non-parametric regardless of n; (3) n $\geq$ 30 $\rightarrow$ parametric via the Central Limit Theorem; (4) otherwise Shapiro-Wilk p > 0.05 $\rightarrow$ parametric. If TRUE, uses Shapiro-Wilk alone (can be over-sensitive at large n). If FALSE, defaults to mean $\pm$ SD for all numeric variables unless specified in force_ordinal.

Logical; if TRUE, includes Shapiro-Wilk P values as an additional column in the output. Default is FALSE.

Logical; if TRUE, rounds all means, medians, IQRs, and standard deviations to nearest integer (0.5 rounds up). Default is FALSE.

Integer or NULL; number of decimal places for all continuous summary values (means, SDs, medians, IQRs). Overrides the default of 1 decimal place when set. Ignored when round_intg = TRUE. Default is NULL (1 decimal place).

Logical; if TRUE, automatically cleans variable names and subheadings for publication-ready output using built-in rule-based pattern matching for common medical abbreviations and prefixes. Default is TRUE.

Logical; if TRUE (default), creates a hierarchical structure with a header row and indented sub-category rows for categorical variables with 3 or more levels. Binary variables (Y/N, YES/NO, or numeric 1/0 – which are auto-detected and treated as Y/N) are always displayed as a single row showing the positive/yes count regardless of this setting. Two-level categorical variables whose values are not Y/N, YES/NO, or 1/0 (e.g. Male/Female) use the hierarchical sub-row format, showing both levels as indented rows. If FALSE, all categorical variables use a single-row flat format. Default is TRUE.

Character; controls the ordering of factor levels in the output. "mixed" (default) applies level-aware ordering for two-level categorical variables and frequency ordering for variables with three or more levels: for any factor, factor level order is always respected regardless of the number of levels; for non-factor two-level variables, levels are sorted alphabetically; for non-factor variables with three or more levels, levels are sorted by decreasing frequency. "levels" respects the original factor level ordering for all variables; if the variable is not a factor, falls back to frequency ordering. "frequency" orders all levels by decreasing frequency (most common first).

Logical; if TRUE (default), generates a methods document describing the statistical presentation used. The document contains boilerplate text for all three table types so the relevant section can be copied directly into a manuscript.

Character; filename for the methods document. Default is "TernTables_methods.docx".

Named character vector specifying where to insert category headers. Names are the header label text to display; values are the anchor variable – either the original column name (e.g. "Age_Years") or the cleaned display name (e.g. "Age (yr)"). Both forms are accepted. Example: c("Demographics" = "Age_Years", "Clinical Measures" = "bmi"). Default is NULL (no category headers).

Named character vector, same interface as category_start. Names are the label text; values are the anchor variable to insert before. Inserts a label-only row with underline formatting and no bold, merge, or border treatments. Default NULL.

Numeric; font size for Word document output tables. Default is 9.

Character vector of display variable names (post-cleaning) that should be formatted as italicized and indented in Word output – matching the appearance of factor sub-category rows. Has no effect on the returned tibble; only applies when output_docx is specified. Default is NULL.

Character vector of display variable names (post-cleaning) that should be formatted as underlined in Word output – matching the appearance of multi-category variable headers. Has no effect on the returned tibble; only applies when output_docx is specified. Default is NULL.

Optional character string for a table caption to display above the table in the Word document. Rendered as size 11 Arial bold, single-spaced with a small gap before the table. Default is NULL (no caption). Example: "Table 1. Patient demographics."

Optional character string for a footnote to display below the table in the Word document. Rendered as size 6 Arial italic with a double-bar border above and below. Default is NULL (no footnote).

Optional character string listing abbreviations. Always printed first in the footnote block. Default NULL.

Optional named character vector. Names are display variable names (case-insensitive); values are the footnote definition text. Each variable gets the next symbol appended to its name in the table, and the footnote block lists each definition below the abbreviation line. To share one footnote between multiple variables, separate their names with a pipe: c("Var A|Var B" = "Shared note text."). Default NULL.

Character; "symbols" (default) uses *, dagger, double-dagger ... "alphabet" uses Unicode superscript letters. See word_export for details.

Logical; if TRUE (default), column headers are wrapped with \n – the first column header includes a category hierarchy label, and the sample size appears on a second line. Set to FALSE to suppress all header line breaks. Can also be set package-wide via options(TernTables.line_break_header = FALSE).

Logical; if TRUE (default), automatically opens the written Word document in the system default application after saving. Set to FALSE to suppress. Has no effect when output_docx is NULL.

Logical; if TRUE (default), appends a citation line at the bottom of the table footnote block and at the end of the methods document: package version, authors, and links to the GitHub repository and web interface. Set to FALSE to suppress.

Character; font family name used for all Word output (table, captions, footnotes, methods document). Any font installed on the system that renders the document may be used. Popular options include "Arial", "Helvetica", "Times New Roman", "Garamond", and "Calibri". Defaults to getOption("TernTables.font_family", "Arial").

Logical; if TRUE, appends a "Missing" row after each variable's data rows showing the count and percentage of missing observations (denominator is the total N). Only emitted when at least one observation is missing. A footnote is automatically appended noting that missing values are reported. Default is FALSE.

Logical; if TRUE, replaces any categorical cell that would display "0 (0%)" with "-" in the output table. Useful when zero counts are not meaningful to report numerically. Default is FALSE.

Controls whether a "Missing, n (%)" column is appended to the table after the Total column. Options:
FALSE (default) — no missingness column added.
"total" — one column appended showing the count and percentage of missing observations across all rows for each variable.
"group" is not supported by ternD() (which has no group structure); use ternG() with show_missingness = "group" instead. Missingness is computed on the raw data column so both NA values and string representations of missing data (e.g., "Unknown", "N/A") are counted. See missing_indicators.

Optional character vector of string values to treat as missing in addition to (or instead of) the built-in ternP defaults. When NULL (default), the ternP canonical list is used. When supplied, the custom list replaces the ternP defaults. Matching is case-insensitive and trims whitespace.

Tibble containing all variables.

Character vector of variables to summarize. Defaults to all except group_var and exclude_vars.

Character vector of variable(s) to exclude. group_var is automatically excluded.

Character, the grouping variable (factor or character with >=2 levels).

Character vector of variables to treat as ordinal (i.e., use medians/IQR and nonparametric tests).

Character vector of variable names to treat as normally distributed, bypassing all normality assessment (Gates 1–4 under "ROBUST", or Shapiro-Wilk under TRUE). Listed variables are summarized as mean $\pm$ SD and compared with Welch tests regardless of the consider_normality setting. Takes priority over consider_normality but not over force_ordinal (if a variable appears in both, force_ordinal wins). Default is NULL.

Character vector of variables to force treatment as continuous (mean $\pm$ SD and parametric tests), bypassing the automatic binary 0/1 detection that would otherwise convert them to categorical Y/N. Useful when a numeric variable with only two unique values (e.g. 0/1 dose levels) should be analysed as a continuous measurement rather than a dichotomous category. Takes priority over automatic type detection but does not override force_ordinal (if a variable appears in both, force_ordinal wins). Default is NULL.

Optional character vector to specify a custom group level order.

Optional filename to export the table as an Excel file.

Optional filename to export the table as a Word document.

Logical; if TRUE, adds unadjusted odds ratios with 95% CI for binary categorical variables (Y/N, YES/NO, or numeric 0/1) and two-level categorical variables (e.g. Male/Female). For two-level categoricals displayed with sub-rows, the reference level (factor level 1, or alphabetical first for non-factors) shows "1.00 (ref.)"; the non-reference level shows the computed OR with 95% CI. Variables with three or more levels show "-". Only valid when group_var has exactly 2 levels; an error is raised for 3+ group comparisons. Default is FALSE.

Character; controls how odds ratios are calculated when OR_col = TRUE. If "dynamic" (default), uses Fisher's exact method when any expected cell count is < 5 (Cochran criterion), otherwise uses the Wald method. If "wald", forces the Wald method regardless of expected cell counts.

Character or logical; controls how continuous variables are routed to parametric vs. non-parametric tests. "ROBUST" (default) applies a four-gate decision consistent with standard biostatistical practice: (1) any group n < 3 is a conservative fail-safe to non-parametric; (2) absolute skewness > 2 or excess kurtosis > 7 in any group routes to non-parametric regardless of sample size (catches heavy-tailed distributions and LOS/count-style skews in which the CLT guarantee is compromised); (3) all groups n $\geq$ 30 routes to parametric via the Central Limit Theorem; (4) otherwise Shapiro-Wilk p > 0.05 in all groups routes to parametric. Normal variables use mean $\pm$ SD and Welch t-test (2 groups) or Welch ANOVA (3+ groups); non-normal variables use median [IQR] and Wilcoxon rank-sum (2 groups) or Kruskal-Wallis (3+ groups). If TRUE, uses Shapiro-Wilk alone (p > 0.05 in all groups = normal). Conservative at large n. If FALSE, all numeric variables are treated as normally distributed regardless of distribution. If "FORCE", all numeric variables are treated as non-normal (median [IQR], nonparametric tests).

Logical; if TRUE, includes Shapiro-Wilk P values in the output. Default is FALSE.

Logical; if TRUE, includes the statistical test name as a column in the output. Default is FALSE.

Integer; number of decimal places for P values (default 3).

Logical; if TRUE, rounds all means, medians, IQRs, and standard deviations to nearest integer (0.5 rounds up). Default is FALSE.

Integer or NULL; number of decimal places for all continuous summary values (means, SDs, medians, IQRs). Overrides the default of 1 decimal place when set. Ignored when round_intg = TRUE. Default is NULL (1 decimal place).

Logical; if TRUE, automatically cleans variable names and subheadings for publication-ready output using built-in rule-based pattern matching for common medical abbreviations and prefixes. Default is TRUE.

Logical; if TRUE (default), creates a hierarchical structure with a header row and indented sub-category rows for categorical variables with 3 or more levels. Binary variables (Y/N, YES/NO, or numeric 1/0 – which are auto-detected and treated as Y/N) are always displayed as a single row showing the positive/yes count regardless of this setting. Two-level categorical variables whose values are not Y/N, YES/NO, or 1/0 (e.g. Male/Female) use the hierarchical sub-row format, showing both levels as indented rows. If FALSE, all categorical variables use a single-row flat format. Default is TRUE.

Character; controls the ordering of factor levels in the output. "mixed" (default) applies level-aware ordering for two-level categorical variables and frequency ordering for variables with three or more levels: for any factor, factor level order is always respected regardless of the number of levels; for non-factor two-level variables (e.g. Male/Female), levels are sorted alphabetically; for non-factor variables with three or more levels, levels are sorted by decreasing frequency. "levels" respects the original factor level ordering for all variables; if the variable is not a factor, falls back to frequency ordering. "frequency" orders all levels by decreasing frequency (most common first).

Numeric; font size for Word document output tables. Default is 9.

Logical; if TRUE (default), generates a methods document describing the statistical tests used.

Character; filename for the methods document. Default is "TernTables_methods.docx".

Named character vector specifying where to insert category headers. Names are the header label text to display; values are the anchor variable – either the original column name (e.g. "Age_Years") or the cleaned display name (e.g. "Age (yr)"). Both forms are accepted. Example: c("Demographics" = "Age_Years", "Clinical" = "bmi"). Default is NULL (no category headers).

Named character vector, same interface as category_start. Names are the label text; values are the anchor variable to insert before. Inserts a label-only row with underline formatting and no bold, merge, or border treatments. Default NULL.

Character vector of display variable names (post-cleaning) that should be formatted as italicized and indented in Word output – matching the appearance of factor sub-category rows. Has no effect on the returned tibble; only applies when output_docx is specified or when the tibble is passed to word_export.

Character vector of display variable names (post-cleaning) that should be formatted as underlined in Word output – matching the appearance of multi-category variable headers. Has no effect on the returned tibble; only applies when output_docx is specified or when the tibble is passed to word_export.

Logical; if FALSE (default), the internal .indent helper column is dropped from the returned tibble. Set to TRUE to retain it – this is necessary when you intend to post-process the tibble and later pass it to word_export directly, as word_export uses the .indent column to apply correct indentation and italic formatting in the Word table.

Logical; if TRUE, adds a "Total" column showing the aggregate summary statistic across all groups (e.g., for a publication Table 1 that includes both per-group and overall columns). Default is TRUE.

Optional character string for a table caption to display above the table in the Word document. Rendered as size 11 Arial bold, single-spaced with a small gap before the table. Default is NULL (no caption). Example: "Table 2. Comparison of recurrence vs. no recurrence."

Optional character string for a footnote to display below the table in the Word document. Rendered as size 6 Arial italic with a double-bar border above and below. Default is NULL (no footnote).

Optional character string listing abbreviations. Always printed first in the footnote block. Default NULL.

Optional named character vector. Names are display variable names (case-insensitive); values are the footnote definition text. Each variable gets the next symbol appended to its name in the table, and the footnote block lists each definition below the abbreviation line. To share one footnote between multiple variables, separate their names with a pipe: c("Var A|Var B" = "Shared note text."). Default NULL.

Character; "symbols" (default) uses *, dagger, double-dagger ... "alphabet" uses Unicode superscript letters. See word_export for details.

Logical; if TRUE (default), column headers are wrapped with \n – group names break on spaces, sample size counts move to a second line, and the first column header reads "Category / Variable". Set to FALSE to suppress all header line breaks. Can also be set package-wide via options(TernTables.line_break_header = FALSE).

Logical; if TRUE, runs pairwise post-hoc tests for continuous and ordinal variables in three or more group comparisons and annotates each group column value with a compact letter display (CLD) superscript. Groups sharing a letter are not significantly different at $\alpha = 0.05$. For normally distributed variables (Welch ANOVA path), Games-Howell pairwise tests are used. For non-normal and ordinal variables (Kruskal-Wallis path), Dunn's test with Holm correction is used. Post-hoc testing is never applied to categorical variables. Only valid when group_var has three or more levels; silently ignored for two-group comparisons. Requires the rstatix package. Default is FALSE.

Logical; if TRUE, applies the Benjamini-Hochberg (BH) false discovery rate correction to all omnibus P values after testing. The correction pool is one P value per variable — sub-rows of multi-level categoricals share the parent P value and are not double-counted. Post-hoc pairwise P values (which already carry within-variable Holm correction) are excluded from the correction pool. Default is FALSE.

Character; controls how BH-corrected P values appear in the output when p_adjust = TRUE. "fdr_only" (default) replaces the standard P value column with the corrected values, renaming the column to "P value (FDR corrected)". "both" retains the original P values in a column named "P value" and adds FDR-corrected values immediately to its right in a column named "P value (FDR corrected)". Ignored when p_adjust = FALSE.

Logical; if TRUE (default), automatically opens the written Word document in the system default application after saving. Set to FALSE to suppress. Has no effect when output_docx is NULL.

Logical; if TRUE (default), appends a citation line at the bottom of the table footnote block and at the end of the methods document: package version, authors, and links to the GitHub repository and web interface. Set to FALSE to suppress.

Character; font family name used for all Word output (table, captions, footnotes, methods document). Any font installed on the system that renders the document may be used. Popular options include "Arial", "Helvetica", "Times New Roman", "Garamond", and "Calibri". Defaults to getOption("TernTables.font_family", "Arial").

Logical; if TRUE, appends a "Missing" row after each variable's data rows showing the count and percentage of missing observations per group (denominator is each group's total N). Missing rows display "-" in the P and OR columns. A footnote is automatically appended noting that missing values are reported. Default is FALSE.

Logical; if TRUE (default), the P value column is included in the output and Excel/Word exports. Set to FALSE to produce a descriptive-only grouped table — the output will contain only the Variable column, one column per group level, and the Total column (if show_total = TRUE). When FALSE, OR_col, show_test, print_normality, post_hoc, categorical_posthoc, and p_adjust are all suppressed automatically.

Logical; if TRUE, replaces any categorical cell that would display "0 (0%)" with "-" in the output table. Useful when zero counts in a group are not meaningful to report numerically (e.g. no patients with a condition in one arm). Default is FALSE.

Character; controls the denominator used when computing percentages for categorical variables. "column" (default) divides each cell count by the column (group) total, so percentages describe the composition of each group – the standard Table 1 interpretation (e.g. "60% of the Recurrence group is Male"). "row" divides each cell count by the row total (the number of subjects with that category level across all groups), so percentages describe how each category level is distributed across groups (e.g. "30% of Males had Recurrence"). When "row", the Total column is automatically suppressed (a Total column would show 100% for every level, which is uninformative). Applies to both binary and multinomial categorical variables in both two- and three-group comparisons.

Logical; if TRUE, computes adjusted standardized residuals from the global contingency table for categorical variables following a significant omnibus test (p < 0.05). Cells whose adjusted standardized residual exceeds $\pm 1.96$ are marked with an asterisk (*), indicating a significant deviation from expected frequencies ($\alpha = 0.05$). This method is equivalent to Haberman's adjusted residuals and does not require a separate multiple-comparisons correction, as the $\pm 1.96$ threshold is derived directly from the omnibus test distribution. Only applied when group_var has three or more levels; silently ignored for two-group comparisons. Default is FALSE.

Controls whether a column of missing-value percentages is appended to the table. Options:
FALSE (default) — no missingness columns added.
"total" — one column ("Missing, n (%)") appended at the far right showing the number and percentage of missing observations across all rows for each variable.
"group" — one column per group level ("Miss. [level]") interleaved immediately after each group's data column, showing per-group missingness for each variable.
Missingness is computed on the raw data column (before ternG's internal NA filtering), so both NA values and string representations of missing data (e.g., "Unknown", "N/A") are counted. See missing_indicators to customise which strings count.

Optional character vector of string values to treat as missing in addition to (or instead of) the built-in ternP defaults. When NULL (default), the ternP canonical list is used ("na", "n/a", "unknown", etc.). When supplied, the custom list replaces the ternP defaults — only the values in missing_indicators (plus true NA) are counted as missing. Matching is always case-insensitive and ignores leading/trailing whitespace.

A data frame or tibble as loaded from a CSV or XLSX file (e.g. via readr::read_csv() or readxl::read_excel()). All character columns are processed; numeric and logical columns are passed through unchanged by the string-cleaning steps.

Preprocessing mode. One of "auto" (default) or "manual".

"auto"

Default behaviour. PHI column-name detection and unnamed-column checks run as hard stops before any cleaning. All other transformations (string-NA conversion, whitespace trimming, empty-column removal, blank-row removal, case normalisation) run automatically.

"manual"

PHI detection is skipped. You take full responsibility for ensuring no patient identifiers are present. A prominent warning is emitted. All other cleaning transformations still run. Use drop_cols to explicitly remove any columns you do not want in the cleaned data.

Optional character vector of additional string values to treat as missing (converted to NA). These are appended to the built-in list (na, n/a, missing, unknown, etc.) — not a replacement. Matching is case-insensitive and whitespace-trimmed. Works in both "auto" and "manual" modes. Example: extra_na = c("9999", "Not Done", "PENDING").

Optional character vector of column names to drop from the data before cleaning begins. Intended for use in "manual" mode to explicitly remove identifier or unwanted columns without triggering the PHI check. Any names not found in data are silently ignored. Works in both modes.

A data frame or tibble. The first column is used as the row-label column (rendered as "Variable" unless renamed via col1_name). All columns are coerced to character before rendering; NA values become empty strings.

Output file path ending in .docx. Pass NULL (default) to write to a temporary file and suppress auto-opening – useful when the result will be passed directly to ternB for bundling.

Optional character string. If supplied, the first column is renamed to this label in the rendered table. The column need not be named "Variable" in the input; any name is accepted and renamed here. Default NULL (use the tibble's existing first column name).

Character vector of labels that already exist as rows in tbl and should be formatted as full section-header rows: cells merged across all columns, bold, with a bottom border line – identical to the treatment applied by category_start. No row is inserted; the matching existing row is formatted in place. Matching is case-insensitive. Default NULL.

Integer vector of body row indices (1-based, final rendered table) to bold across every column. Applied after all structural formatting so it always wins. Default NULL.

Optional named list for cell-level p-value-based bolding. Use this when your tibble has pre-formatted p-value strings in columns that are not named "P value" (e.g. "Uni p", "Multi p"). Supply a list with:

• p_cols — character vector of column names containing p-value strings.

• hr_cols — optional character vector of column names (same length and order as p_cols) to also bold when the paired p-value is significant (e.g. the corresponding HR or coefficient column). Pass NULL or omit to skip paired-column bolding.

• threshold — numeric significance threshold. Default 0.05.

The Variable column is never modified by bold_sig; use bold_rows to bold entire rows (e.g., predictor-level rows where the p-value represents an omnibus LRT). Example:

bold_sig = list(
  p_cols    = c("Uni p", "Multi p"),
  hr_cols   = c("Uni HR (95% CI)", "Multi HR (95% CI)"),
  threshold = 0.05
)

Default NULL.

Integer vector of body row indices to italicize across every column. Default NULL.

Integer vector of column indices (1-based) to bold across all body rows. Default NULL.

Integer vector of column indices to italicize across all body rows. Default NULL.

Logical; if TRUE, columns listed in bold_cols or italic_cols also have their header cell bolded or italicized. Default FALSE.

Logical; passed to word_export. Default FALSE.

Integer or NULL; if provided, rounds all numeric values in the table to this many decimal places before rendering. Passed to word_export. Default NULL (no rounding).

Numeric; font size for table body. Default 9.

Named character vector; same as in word_export. Insert new section-header rows at anchor variable positions, in addition to any rows already in the tibble. Default NULL.

Named character vector; same as in word_export. Insert underline-only (no bold, no merge) label rows at anchor positions. Default NULL.

Character vector of row labels to italicize and indent (sub-item appearance). Default NULL.

Character vector of row labels to underline (multi- category header appearance without the full subheader treatment). Default NULL.

Optional character string for the caption above the table. Default NULL.

Optional character string for a footnote below the table. Default NULL.

Optional character string (or character vector) of abbreviations. Always printed first in the footnote block. Default NULL.

Optional named character vector of per-variable footnote definitions (case-insensitive name match). To share one footnote symbol between multiple variables, separate their names with a pipe: c("Var A|Var B" = "Shared note text."). Default NULL.

Character; "symbols" (default) or "alphabet". Controls the footnote symbol sequence. See word_export for details.

Optional character string. Overrides the top-left header cell. When NULL (default), the standard "Category\n Variable" label is used. Example: "Variable\n Index Management Strategy".

Logical; if TRUE, column headers are wrapped with \n and the first column header shows the two-line "Category / Variable" label. For custom tibbles the column names are typically already formatted, so this defaults to FALSE here (unlike word_export where it defaults to TRUE).

Logical; if TRUE (default), opens the written document after saving. Default TRUE.

Logical; if TRUE (default), appends the TernTables citation line in the page footer. Default TRUE.

Character; font family name used for all Word output. Defaults to getOption("TernTables.font_family", "Arial"). See word_export for details.

Numeric mean value. Formatted to 1 decimal place.

Numeric standard deviation. Formatted to 1 decimal place.

Numeric P value in the range [0, 1]. NA values are returned as NA_character_. Values >= 1 (or rounding to >= 1) are returned as e.g. ">0.999".

Integer; number of decimal places for reported P values. Default is 3. Note: for p < 0.001, the value is reported in scientific notation with 1 significant figure regardless of digits (e.g., 8E-4).

A tibble created by ternG or ternD

Output file path ending in .docx

Logical; if TRUE, adds note about integer rounding. Default is FALSE.

Integer or NULL; if provided, rounds all numeric values in the table to this many decimal places before rendering. Default is NULL (no rounding).

Numeric; font size for table body. Default is 9.

Named character vector specifying category headers. Names are header label text; values are anchor variable names – either the original column name or the cleaned display name (both forms accepted).

Named character vector, same interface as category_start. Names are the label text to display; values are the anchor variable to insert before. The inserted row has text only in column 1 (all other cells blank) and receives underline formatting – identical to manual_underline – but no bold, merge, or border treatments. Default is NULL (none).

Character vector of labels that already exist as rows in the table and should be formatted as full category section headers (merged across all columns, bold, with a bottom border line). Unlike category_start, no new row is inserted – the matching existing row is formatted in place. Intended for use with ternStyle() where section- divider rows are pre-built into the tibble. Case-insensitive match. Default NULL.

Integer vector of body row indices (1-based, in the final rendered table) to bold across every column. Applied as the last formatting pass so it overrides structural formatting. Default NULL.

Optional named list for cell-level conditional bolding based on parsed p-values. Intended for use with ternStyle() when columns contain pre-formatted p-values that are not named "P value" (the name ternG uses internally). Supply a list with:

• p_cols — character vector of column names containing p-value strings.

• hr_cols — optional character vector of column names to also bold when the paired p-value is significant (must be the same length and order as p_cols, or NULL to skip paired HR bolding).

• threshold — numeric significance threshold; default 0.05.

For each p-value cell where the parsed numeric value is below threshold, that cell is bolded. If hr_cols is supplied, the corresponding HR cell in the same row is also bolded. The Variable column is never touched by this argument — use bold_rows to bold entire rows (e.g., predictor header rows). Default NULL.

Integer vector of body row indices to italicize across every column. Default NULL.

Integer vector of column indices (1-based) to bold across all body rows. Default NULL.

Integer vector of column indices to italicize across all body rows. Default NULL.

Logical; if TRUE, any columns listed in bold_cols or italic_cols also have their header cell bolded or italicized, respectively. Default FALSE.

Character vector of display variable names (post-cleaning) to force into italicized and indented formatting, matching the appearance of factor sub-category rows (e.g., levels of a multi-category variable). Use this for rows that should visually appear as sub-items but are not automatically detected as such.

Character vector of display variable names (post-cleaning) to force into underlined formatting, matching the appearance of multi-category variable header rows. Use this for rows that should visually appear as section headers but are not automatically detected as such.

Optional character string to display as a caption above the table in the Word document. Rendered as size 11 Arial bold, single-spaced with a small gap before the table. Default is NULL (no caption).

Optional character string to display as a footnote below the table in the Word document. Rendered as size 6 Arial italic. A double-bar border is applied above and below the footnote row. Default is NULL (no footnote).

Optional character string (or character vector, which will be collapsed with spaces) listing abbreviations to display at the top of the footnote block. Always printed first, before any variable-specific footnote lines. Default NULL.

Optional character string describing post-hoc CLD superscript conventions. When supplied by ternG(), it is inserted after the abbreviations footnote and before the variable symbol footnotes. Default NULL.

Optional named character vector. Names are display variable names as they appear in the table (case-insensitive match); values are the footnote definition text for that variable. Each entry is assigned the next symbol in the sequence (*, dagger, double-dagger, ...) and the symbol is appended to the variable name in column 1. The footnote block lists each as "* Definition text." below the abbreviations. To map multiple variables to the same footnote symbol and note, separate the variable names with a pipe character in the key: c("Var A|Var B" = "Shared note text."). Default NULL.

Character; controls the footnote symbol sequence. "symbols" (default) uses *, dagger, double-dagger, section, pilcrow, double-vertical-bar, then doubled forms. "*" is appended as plain text; all others are rendered as true Word superscripts. "alphabet" uses Unicode superscript letters (a, b, c, ...) which render as raised glyphs without explicit superscript formatting.

Logical; if TRUE, a page break is appended at the end of the Word document after the table. Used internally by ternB() to embed page breaks inside each table's temp file rather than injecting them into the combined document body, which avoids double-break artifacts when tables do not fill the page. Default is FALSE.

Optional character string. Overrides the top-left header cell text. When NULL (default), the cell shows "Category\n Variable" (the standard two-line label). Supply any string, including "\n" line breaks, to customise. Example: "Variable\n Index Management Strategy".

Logical; if TRUE (default), column headers are wrapped with \n – group names break on spaces, sample size counts move to a second line, and the first column header includes a category hierarchy label. Set to FALSE to suppress all header line breaks. Can also be set package-wide via options(TernTables.line_break_header = FALSE).

Logical; if TRUE (default), automatically opens the written Word document in the system default application after saving. Set to FALSE to suppress.

Logical; if TRUE (default), appends a citation line at the bottom of the table footnote block: package version, authors, and links to the GitHub repository and web interface. Set to FALSE to suppress.

Character; font family used for the entire Word table and its caption, footnote, and citation. Any font name accepted by the rendering system is valid (Word will fall back to its default if the font is not installed). Can also be set package-wide via options(TernTables.font_family = "Garamond"). Default is "Arial".

A ternP_result object returned by ternP.

Output file path ending in .docx. Default is "cleaning_summary.docx" in the current working directory.

Character; font family for the Word document. Default "Arial". Can also be set via options(TernTables.font_family = ...).

Logical; if TRUE (default), automatically opens the written Word document in the system default application after saving. Set to FALSE to suppress.

Logical; if TRUE (default), appends the TernTables citation as a page footer in the written document. Set to FALSE to suppress.

A tibble created by ternG or ternD, or NULL when generating a generic document.

Output file path ending in .docx.

Number of group levels used in ternG (2 for two-group, 3+ for multi-group). Ignored when called from ternD.

Logical; whether odds ratios were calculated. Default FALSE.

Character; the OR calculation method used in ternG. "dynamic" (default) means Wald when all expected cells >= 5, Fisher's exact otherwise. "wald" means Wald was forced regardless of cell counts. Controls the OR description in the generated methods paragraph.

Character; "ternG" or "ternD". Controls which section is populated with dynamic test information. Default "ternG".

Logical; whether pairwise post-hoc testing was requested (post_hoc = TRUE in ternG). When TRUE and n_levels >= 3, the three-group methods paragraph is updated to describe the post-hoc test pairing (Games-Howell or Dunn's + Holm). Default FALSE.

Logical; whether adjusted standardized residuals were requested (categorical_posthoc = TRUE in ternG). When TRUE and n_levels >= 3, the methods paragraph notes that cells with adjusted standardized residuals exceeding $\pm 1.96$ are marked with an asterisk following a significant omnibus test. Default FALSE.

Character vector of variable names for which Fisher's exact test was the omnibus test while categorical_posthoc = TRUE. When non-empty, a caveat sentence is appended noting that Haberman's adjusted residuals were derived from the chi-squared contingency table in the absence of a Fisher's exact equivalent. Populated automatically when called from ternG(). Default character(0).

Logical or character; whether missingness columns were added to the table (FALSE, "total", or "group"). When non-FALSE, a sentence is appended describing the missingness reporting approach and the string representations used to flag missing values. Should match the show_missingness argument passed to ternG() or ternD(). Default FALSE.

Character vector of string values treated as missing in addition to R NA, or NULL to use TernTables defaults. Should match the missing_indicators argument passed to ternG() or ternD(). Default NULL.

Logical; if TRUE, ignores all other arguments and writes a single comprehensive Word document covering every possible TernTables configuration (descriptive, two-group with and without odds ratios, three-or-more-group with and without post-hoc testing), using package-default phrasing throughout. Output is written to filename if supplied; otherwise falls back to comprehensive_boilerplate_methods.docx in the current working directory. Intended as a reference document, not for inclusion in a manuscript. Default FALSE.

Logical; if TRUE, prepends a sentence to the methods paragraph stating that P values were corrected using the Benjamini-Hochberg FDR procedure, and updates the significance threshold wording accordingly. Should match the p_adjust argument passed to ternG(). Default FALSE.

Logical; if TRUE (default), automatically opens the written Word document in the system default application after saving. Set to FALSE to suppress.

Logical; if TRUE (default), appends a citation line after the document footer: package version, authors, and links to the GitHub repository and web interface. Set to FALSE to suppress.

Character; font family for the Word document. Default "Arial". Can also be set via options(TernTables.font_family = ...).