GPCERF
nsaph-software.github.io
- Free plan
- Self-hosted
6.9 Club score Visit nsaph-software.github.io
What the maker publishes
- Bayesian analysis
- Yes source
- Bayesian framework
- Its framework is nonparametric and Bayesian, based on Gaussian process priors. source
- Covariate adjustment
- The documented workflow converts covariate values into a composite generalized propensity score before fitting Gaussian processes. source
- Covariate processing
- The standard Gaussian process workflow converts covariate values into a composite score called the generalized propensity score before fitting Gaussian processes. source
- Curve estimation
- The standard Gaussian process workflow estimates an exposure-response function and visualizes a pointwise 95% credible band. source
- Deployment
- desktop source
- Free plan
- Yes source
- Installation
- The repository README shows installation from GitHub with devtools using the develop branch. source
- Integration
- The documented examples use CausalGPS for synthetic data generation and GPS estimation, and ggplot2 for plotting. source
- Intended users
- The documented workflows estimate causal exposure-response functions from observational or simulated data. source
- License
- The project site lists the license as GPL (>= 3). source
- Maintenance status
- CRAN says GPCERF was removed on 2026-01-12 because issues were not corrected despite reminders. source
- Methods
- The package provides Gaussian process and nearest-neighbor Gaussian process approaches for nonparametric modeling. source
- Outputs
- The examples estimate exposure-response functions and plot pointwise 95% credible bands. source
- Purpose
- GPCERF estimates causal effects of a continuous exposure and detects change points in causal exposure-response curves using observational data. source
- R dependencies
- The repository installation example uses the R devtools package, and its usage note says examples also require the ranger R package. source
- Repository status
- The CRAN page says GPCERF was archived on 2026-01-12 because issues were not corrected despite reminders. source
- Runtime dependency
- The README notes that its usage examples also require the ranger R package. source
- Scaling option
- The nearest-neighbor Gaussian process example reports much shorter running time than the standard Gaussian process example. source
- Scripting support
- Yes source
- Support
- The README directs users to report issues on GitHub and says they can seek assistance through GitHub Discussions or issues. source
Checked against the maker's pages on 12 Oct 2026.