Compute correlations from event data — compute

Computes correlations between test statistics using the mathematically correct formulation that ensures positive definiteness and proper asymptotic properties.

Usage

compute_correlations(event, check = TRUE, return_matrix = TRUE)

Arguments

event: A data frame with columns H1, H2, Analysis, and Event containing event count data for correlation computation
check: Logical indicating whether to perform input validation (default: TRUE)
return_matrix: Logical indicating whether to return as matrix (TRUE) or data frame (FALSE). Default: TRUE

Value

If return_matrix=TRUE, returns a symmetric correlation matrix. If return_matrix=FALSE, returns a data frame with columns H1, H2, Analysis1, Analysis2, Correlation.

Examples

library(tibble)

# Sample event data
event_data <- tribble(
  ~H1, ~H2, ~Analysis, ~Event,
  1, 1, 1, 80,
  2, 2, 1, 100,
  1, 2, 1, 60,
  1, 1, 2, 120,
  2, 2, 2, 150,
  1, 2, 2, 80
)

# Get correlation matrix
corr_matrix <- compute_correlations(event_data)

# Get detailed data frame
corr_df <- compute_correlations(event_data, return_matrix = FALSE)