Plot correlations heatmap

Calculate correlations (or partial correlations if covariates are supplied) between variables and plot them as a heatmap with significance stars.

Usage

PlotCorrelationsHeatmap(
  Data,
  xVars = NULL,
  yVars = NULL,
  covars = NULL,
  method = "pearson",
  Relabel = TRUE,
  Ordinal = FALSE,
  min_n = 3,
  eps = 1e-12
)

Arguments

Data

A data frame containing variables to correlate.

xVars

Character vector of x-axis variable names. If NULL, numeric variables are auto-selected.

yVars

Character vector of y-axis variable names. If NULL, uses xVars and removes diagonal.

covars

Character vector of covariate names. If provided, computes partial correlations via ppcor.

method

Correlation method: "pearson", "spearman", or "kendall".

Relabel

Logical. If TRUE, uses sjlabelled variable labels when present.

Ordinal

Logical. If TRUE, ordinal variables are converted to numeric using ConvertOrdinalToNumeric().

min_n

Minimum complete cases required for a correlation (or partial correlation).

eps

Small tolerance to treat near-constant variables as constant.

Value

A list with:

• Unadjusted: list(r, p, npairs, plot)

• FDRCorrected: list(r, p, npairs, plot)

• method, Relabel, Covariates, CovariatesMissing

Details

This version is robust to non-syntactic column names (spaces, hyphens, Greek letters like β), by preserving names during data.frame coercions and using tidyselect::all_of() in pivot_longer().