Calculate correlations between a target variable and all other numeric variables.

This function computes Pearson or Spearman correlation coefficients between a specified target variable and all other numeric variables in the input data frame. It also calculates p-values and adjusted p-values (Bonferroni and Benjamini-Hochberg). Optionally, it can perform group-wise correlation analysis and compare correlations between two groups using Fisher's z-transformation.

Usage

cyt_corr(
  data,
  target,
  methods = c("spearman", "pearson"),
  group_var = NULL,
  compare_groups = FALSE,
  plot = FALSE,
  progress = NULL
)

Arguments

data: A data frame containing the variables.
target: A character string specifying the name of the target variable. Must be a numeric column in data.
methods: A character string indicating the correlation method to be used. Can be "pearson" (default) or "spearman".
group_var: An optional character string specifying the name of a grouping variable in data. If provided, group-wise correlations will be calculated.
compare_groups: Logical. If TRUE and group_var is provided with at least two levels, the function will compare the correlations between the first two levels of group_var.
plot: Logical. If TRUE, the function will generate and return a correlation plot.
progress: Optional. A Shiny Progress object for reporting progress updates.

Value

A list containing:

results: A data frame with overall correlation results, including variable names, correlation coefficients (r), p-values (p), sample size (n), method, and adjusted p-values (p_bonf, p_bh), sorted by absolute correlation coefficient in descending order.
results_groupwise: A data frame with group-wise correlation results, if group_var is provided. Includes similar columns as results, plus the group identifier.
results_diff: A data frame with results of correlation comparison between the first two groups, if compare_groups is TRUE and group_var has at least two levels. Includes variable, r_diff (difference in correlations), z (Fisher's z-score), p_diff (p-value for the difference), and adjusted p-values (p_diff_bonf, p_diff_bh).
plot: A correlation plot, if plot is TRUE.