Package index • svyder

Core pipeline

The compute-classify-correct workflow. Use der_diagnose() for the all-in-one call, or invoke the three steps individually for finer control.

der_diagnose(): All-in-One DER Diagnostic
der_compute(): Compute Design Effect Ratios
der_classify(): Classify Parameters by Design Sensitivity
der_correct(): Apply Selective Correction to Flagged Parameters

Analysis

Understand, verify, and re-target DER results.

der_decompose(): Decompose DER into Components
der_sensitivity(): Sensitivity Analysis Across Threshold Values
der_theorem_check(): Verify Theoretical DER Predictions
der_compare(): Compare DER Across Variance Targets

Object methods

S3 methods for svyder objects: printing, summaries, corrected draws, and broom-style tidiers.

print(<svyder>): Print a svyder Object
summary(<svyder>): Summarize a svyder Object
tidy(<svyder>): Tidy a svyder Object
glance(<svyder>): Glance at a svyder Object
as.matrix(<svyder>): Extract Draws Matrix from a svyder Object
is.svyder(): Test if an Object is a svyder Object

Visualization

Diagnostic plots for DER results: profile, decomposition, and comparison views, with an automatic ggplot2 / base R fallback.

plot(<svyder>): Plot DER Diagnostic Results
autoplot.svyder(): Create a ggplot2 Visualization of DER Results

Backends

Generic extractors for posterior draws and survey design objects. Methods exist for brms, cmdstanr, rstanarm, posterior, and the survey package.

extract_draws(): Extract Posterior Draws from Model Objects
extract_design(): Extract Survey Design Information

Bundled datasets

Synthetic datasets shipping pre-computed posterior draws, so no Stan installation is required.

nsece_demo: Synthetic NSECE-Like Survey Data
sim_hlr: Simulated Hierarchical Linear Regression Data

Package

svyder svyder-package: svyder: Design Effect Ratio Diagnostics for Bayesian Survey Models