Skip to contents

Standard 4-Quadrant Clinical Contingency Report

Source: R/plot_pairwise_categorical.R
cbe_four_quadrant_report.Rd

Generates a standard 4-quadrant clinical report for a 2x2 contingency table:


q1 | q2
q3 | q4
p = pformat

where q1 is (Row 1, Col 1), q2 is (Row 1, Col 2), q3 is (Row 2, Col 1), and q4 is (Row 2, Col 2). If any cell count is zero, hypothesis testing automatically defaults to the mid-p version of Central Fisher's exact test via exact2x2.

Usage

cbe_four_quadrant_report(
  data,
  var1 = NULL,
  var2 = NULL,
  label1 = NULL,
  label2 = NULL,
  test = c("auto", "exact", "chisq", "fisher"),
  correct = FALSE,
  p_format = NULL,
  ...
)

Arguments

data

A data frame or a 2x2 matrix/table.

var1

Character string of row variable column name (if data is a data frame).

var2

Character string of column variable column name (if data is a data frame).

label1

Display label for row variable.

label2

Display label for column variable.

test

Hypothesis test engine: "auto" (default CBE hierarchy: mid-p exact if zero-cell, otherwise Central Fisher's exact), "exact" (force exact test), "chisq" (force Pearson's Chi-squared test via chisq.test), or "fisher" (force fisher.test).

correct

Logical; whether to apply continuity correction for "chisq" (default FALSE).

p_format

Function for p-value formatting (default pformat).

...

Additional arguments passed to cbe_contingency_plot.

Value

A list containing:

quadrants

A tibble with counts and percentages for q1, q2, q3, and q4.

p_value

Numeric exact test p-value.

p_formatted

Formatted p-value string (e.g. "p = 0.024").

test_method

Name of the exact test used.

compact_report

Compact 3-line string formatted as q1 | q2 // q3 | q4 // p = pformat.

text_card

Console-ready formatted 4-quadrant text card.

plot

A ggplot object rendering the 4-quadrant square report.

Examples

tab <- matrix(c(14, 6, 4, 16), nrow = 2,
              dimnames = list(c("Yes", "No"), c("Active", "Placebo")))
rep <- cbe_four_quadrant_report(tab, label1 = "Response", label2 = "Treatment")
cat(rep$compact_report, "\n\n")
#> q1: n = 14 (35.0%) | q2: n = 4 (10.0%)
#> q3: n = 6 (15.0%) | q4: n = 16 (40.0%)
#> p = 0.004 
#> 
cat(rep$text_card)
#> q1: Yes / Active | q2: Yes / Placebo
#>   n = 14 (35.0%) |   n = 4 (10.0%)
#> -----------------------+-----------------------
#> q3: No / Active | q4: No / Placebo
#>   n = 6 (15.0%) |   n = 16 (40.0%)
#> p = 0.004 (Central Fisher's Exact Test)

# Force Pearson Chi-squared test:
rep_chi <- cbe_four_quadrant_report(tab, test = "chisq")
cat(rep_chi$compact_report, "\n")
#> q1: n = 14 (35.0%) | q2: n = 4 (10.0%)
#> q3: n = 6 (15.0%) | q4: n = 16 (40.0%)
#> p = 0.001