Bayesian Model Averaging for survival analysis
Usage
bayesianma(
data,
time_var,
event_var,
pred_vars,
prior_type = "uniform",
prior_inclusion_prob = 0.5,
beta_alpha = 1,
beta_beta = 1,
complexity_penalty = 1,
mcmc_method = "mc3",
mcmc_chains = 3,
mcmc_iterations = 5000,
burn_in = 1000,
thinning = 1,
convergence_diagnostic = "gelman_rubin",
model_selection_method = "highest_posterior",
occam_ratio = 20,
temperature_ladder = "1.0,1.5,2.0,3.0",
proposal_variance = 1,
variable_selection_threshold = 0.5,
uncertainty_quantification = TRUE,
prediction_intervals = TRUE,
credible_level = 0.95,
model_diagnostics = TRUE,
sensitivity_analysis = FALSE,
cross_validation = FALSE,
cv_folds = 5,
parallel_processing = FALSE,
seed_value = 42,
show_summary = TRUE,
show_model_space = TRUE,
show_coefficients = TRUE,
show_inclusion_probs = TRUE,
show_top_models = TRUE,
show_diagnostics = TRUE,
show_predictions = FALSE,
plot_model_probs = TRUE,
plot_inclusion_probs = TRUE,
plot_coefficients = TRUE,
plot_convergence = TRUE,
plot_model_space = FALSE,
plot_predictions = FALSE,
showExplanations = TRUE
)Arguments
- data
the data as a data frame
- time_var
the time-to-event variable
- event_var
the event indicator variable (0/1 or FALSE/TRUE)
- pred_vars
the predictor variables for model averaging
- prior_type
type of prior distribution for model space
- prior_inclusion_prob
prior probability of variable inclusion (for uniform/beta-binomial priors)
- beta_alpha
alpha parameter for beta-binomial prior
- beta_beta
beta parameter for beta-binomial prior
- complexity_penalty
penalty parameter for complexity prior
- mcmc_method
MCMC sampling method for model space exploration
- mcmc_chains
number of MCMC chains
- mcmc_iterations
number of MCMC iterations per chain
- burn_in
number of burn-in iterations
- thinning
thinning interval for MCMC samples
- convergence_diagnostic
convergence diagnostic method
- model_selection_method
method for selecting representative model
- occam_ratio
ratio for Occam's window model selection
- temperature_ladder
temperature ladder for MC³ (comma-separated values). Higher temperatures enable better mixing
- proposal_variance
variance for parameter proposal distributions
- variable_selection_threshold
posterior probability threshold for variable selection
- uncertainty_quantification
whether to perform comprehensive uncertainty quantification
- prediction_intervals
whether to compute prediction intervals
- credible_level
level for credible intervals
- model_diagnostics
whether to perform comprehensive model diagnostics
- sensitivity_analysis
whether to perform prior sensitivity analysis
- cross_validation
whether to perform cross-validation assessment
- cv_folds
number of folds for cross-validation
- parallel_processing
whether to use parallel processing for MCMC
- seed_value
random seed for reproducible results
- show_summary
show model averaging summary
- show_model_space
show model space exploration results
- show_coefficients
show model-averaged coefficient estimates
- show_inclusion_probs
show posterior inclusion probabilities
- show_top_models
show highest probability models
- show_diagnostics
show MCMC convergence diagnostics
- show_predictions
show model-averaged predictions
- plot_model_probs
plot model posterior probabilities
- plot_inclusion_probs
plot variable inclusion probabilities
- plot_coefficients
plot posterior coefficient distributions
- plot_convergence
plot MCMC convergence diagnostics
- plot_model_space
plot model space exploration
- plot_predictions
plot model-averaged predictions
- showExplanations
show explanations for the analysis
Value
A results object containing:
results$instructions | a html | ||||
results$todo | a html | ||||
results$summary | a table | ||||
results$modelSpace | a table | ||||
results$averagedCoefficients | a table | ||||
results$inclusionProbabilities | a table | ||||
results$topModels | a table | ||||
results$mcmcDiagnostics | a table | ||||
results$selectedModel | a table | ||||
results$uncertaintyQuantification | a table | ||||
results$sensitivityAnalysis | a table | ||||
results$crossValidation | a table | ||||
results$modelProbsPlot | an image | ||||
results$inclusionProbsPlot | an image | ||||
results$coefficientsPlot | an image | ||||
results$convergencePlot | an image | ||||
results$modelSpacePlot | an image | ||||
results$predictionsPlot | an image | ||||
results$explanations | a html |
Tables can be converted to data frames with asDF or as.data.frame. For example:
results$summary$asDF
as.data.frame(results$summary)