qbrms (version 1.0.1)

diagnose_model: Automated Model Diagnostics and Recommendations

Description

Comprehensive automated diagnostics for qbrms models with actionable recommendations for model improvement.

Usage

diagnose_model(model, checks = "all", verbose = TRUE)

Value

An object of class "qbrms_diagnostics" containing:

• summary: Overall assessment (pass/warning/fail)

• checks: Detailed results for each diagnostic check

• recommendations: Specific suggestions for improvement

• plots: List of diagnostic plots

Arguments

model

A fitted qbrms model object

checks

Character vector specifying which checks to perform. Options: "all" (default), "convergence", "fit", "residuals", "posterior", "influential"

verbose

Logical; if TRUE, prints detailed diagnostic information (default: TRUE)

Details

This function performs comprehensive model diagnostics including:

• Convergence checks (for MCMC-based inference)

• Goodness-of-fit assessment

• Residual analysis

• Posterior predictive checks

• Influential observation detection

• Prior-posterior overlap assessment

Each check produces a pass/warning/fail status with specific recommendations for addressing any issues detected.

Examples

if (FALSE) {
# Fit a model
fit <- qbrms(mpg ~ hp + wt, data = mtcars, family = gaussian())

# Run diagnostics
diag <- diagnose_model(fit)

# View summary
print(diag)

# View specific recommendations
diag$recommendations

# Create diagnostic plots
plot(diag)
}