# R overview The `survivalGPU` R package performs survival analysis on GPU-accelerated hardware. Currently, two models are implemented: - **`coxphGPU()`** — Cox Proportional Hazards, built to take the same inputs as `survival::coxph()`. - **`wceGPU()`** — Weighted Cumulative Exposure (WCE), built to take the same inputs as `WCE::WCE()`. Both support bootstrap resampling, as a separate step after the fit (`bootstrap()`), and both run on CPU if no CUDA-capable GPU is available. Under the hood, the R package delegates its computations to the same Python backend as the `survivalgpu` Python package, via [reticulate](https://rstudio.github.io/reticulate/) — see {doc}`installation` to set that up. For full function-level documentation, see the R API reference (built with [pkgdown](https://pkgdown.r-lib.org/)).