evanverse (version 0.4.0)

quick_anova: Quick ANOVA with Automatic Method Selection

Description

Conduct one-way ANOVA, Welch ANOVA, or Kruskal-Wallis test with automatic assumption checks, publication-ready visualization, and optional post-hoc comparisons. Designed for comparing two or more independent groups.

Usage

quick_anova(
  data,
  group,
  value,
  method = c("auto", "anova", "welch", "kruskal"),
  post_hoc = c("auto", "none", "tukey", "welch", "wilcox"),
  conf.level = 0.95,
  plot_type = c("boxplot", "violin", "both"),
  add_jitter = TRUE,
  point_size = 2,
  point_alpha = 0.6,
  show_p_value = TRUE,
  p_label = c("p.format", "p.signif"),
  palette = "qual_vivid",
  verbose = TRUE,
  ...
)

Value

An object of class quick_anova_result with elements:

plot

ggplot object of the comparison

omnibus_result

List describing the main test

post_hoc

Post-hoc comparison table (if requested)

method_used

Character. "anova", "welch", or "kruskal"

descriptive_stats

Summary statistics by group

assumption_checks

Results of normality/variance checks

auto_decision

Details explaining automatic selections

timestamp

POSIXct timestamp of the analysis

Arguments

data

A data frame containing the variables.

group

Column name for the grouping factor. Supports quoted or unquoted names via tidy evaluation.

value

Column name for the numeric response variable.

method

Character. One of "auto" (default), "anova", "welch", or "kruskal". When "auto", the function inspects normality and homogeneity of variances to pick an appropriate test.

post_hoc

Character. Post-hoc procedure: "auto" (default), "none", "tukey", "welch", or "wilcox". "auto" selects Tukey for ANOVA, Welch-style pairwise t-tests for Welch ANOVA, and pairwise Wilcoxon tests for Kruskal-Wallis.

conf.level

Numeric. Confidence level for the test/intervals. Default is 0.95.

plot_type

Character. One of "boxplot", "violin", or "both".

add_jitter

Logical. Add jittered points? Default TRUE.

point_size

Numeric. Size of jitter points. Default 2.

point_alpha

Numeric. Transparency for jitter points (0-1). Default 0.6.

show_p_value

Logical. Show omnibus p-value on the plot? Default TRUE.

p_label

Character. P-value display: "p.format" (default) or "p.signif" (stars).

palette

Character. Palette name from evanverse, or NULL for defaults.

verbose

Logical. Print informative messages? Default TRUE.

...

Reserved for future extensions.

See Also

aov, oneway.test, kruskal.test

Examples

set.seed(123)
df <- data.frame(
  group = rep(LETTERS[1:3], each = 40),
  value = rnorm(120, mean = rep(c(0, 0.5, 1.2), each = 40), sd = 1)
)
res <- quick_anova(df, group, value)
res$plot
summary(res)