ETA-covariate correlation table

Description

Computes Pearson correlations between empirical Bayes estimates (ETAs) and covariates in the original dataset. Identifies which covariates are most worth testing in a formal covariate search. Only columns that are constant within each subject are treated as covariates.

Usage

ferx_eta_cov(fit, data)

Arguments

fit: A ferx_fit object returned by [ferx_fit](ferx_fit.qmd).
data: The original dataset (data frame) passed to [ferx_fit](ferx_fit.qmd).

Other diagnostics: [check_diagnostics](check_diagnostics.qmd)(), [ferx_cor_matrix](ferx_cor_matrix.qmd)(), [ferx_cov_screen](ferx_cov_screen.qmd)(), [ferx_estimates](ferx_estimates.qmd)(), [ferx_plot_trace](ferx_plot_trace.qmd)(), [ferx_runlog_iters](ferx_runlog_iters.qmd)(), [ferx_trace](ferx_trace.qmd)(), [ferx_warnings](ferx_warnings.qmd)(), [summary.ferx_fit](summary.ferx_fit.qmd)()

Concept

diagnostics

Value

Data frame with columns eta, covariate, r, p_val, flag, sorted by descending |r|. Returned invisibly; the full table is printed to the console.

Examples

ex  <- ferx_example("warfarin")
fit <- ferx_fit(ex$model, ex$data, method = "gn", covariance = FALSE)
obs <- read.csv(ex$data)
ferx_eta_cov(fit, obs)