Skip to contents

Tidy Proportional Hazards Diagnostics for Cox Models

Source: R/cbe_cox_check.R
cox_check.Rd

Standalone proportional hazards (PH) assumption diagnostics, usable on any fitted survival::coxph model, or on a cbe_cox or cbe_cox_multi object. Wraps survival::cox.zph() with a tidy per-term table, a per-term violation flag, and an automated text summary.

Usage

cox_check(fit)

Arguments

fit

A fitted survival::coxph model, or a cbe_cox or cbe_cox_multi object.

Value

An object of class cbe_cox_check containing:

  • zph: The survival::cox.zph object (or NULL if it could not be computed).

  • zph_table: Data frame version of zph$table (one row per term, plus a GLOBAL row for multivariable fits).

  • zph_violated: Named logical vector, one entry per term (excluding GLOBAL), TRUE where the assumption is violated (p < 0.05). For a univariable fit this is a length-1 named logical vector.

  • zph_text: Automated summary sentence (a single string for a univariable fit with one term; a two-element character vector, per-term then global, for a multivariable fit).

See also

[cox_single()], [cox_multi()]