Package index • scdbayes

Getting Started

Launch the Shiny app

Identify design, family, hierarchy, and column roles

detect_design_type(): Detect SCD design type
detect_hierarchical_structure(): Detect hierarchical structure in data
detect_hierarchy_levels(): Detect hierarchy levels in SCD data
find_case_var(): Locate the case-level grouping column by common SCED name patterns
find_cluster_var(): Locate a mid-level cluster grouping column
find_nest_var(): Detect a within-case nesting variable by name pattern then structure
find_phase_var(): Locate the phase column by common SCED name patterns
find_study_var(): Locate the study- / institution-level grouping column
find_time_var(): Locate the time column by common SCED name patterns
recommend_distribution_family(): Recommend distribution family based on outcome data
validate_scd_data(): Validate SCD data

Process phases, compute case-level features, generate synthetic data

process_scd_phases(): Build piecewise regression coding for single-case design phases
consolidate_phase_indicators(): Consolidate binary phase indicator columns into a single phase column
compute_autocorrelation_features(): Lag-1 autocorrelation per phase per case
compute_case_features(): Per-case descriptive features
analyze_covariates(): Analyze continuous and categorical covariates
analyze_design_patterns(): Analyze design patterns
simulate_scd_data(): Simulate SCD data

Bayesian piecewise model formulas, priors, and fits via brms

build_default_random_effects(): Build default random effects formula string
build_comparison_models(): Build candidate model formulas
generate_adaptive_priors(): Generate adaptive priors based on data characteristics
generate_prior_recommendations(): Generate prior recommendations
assemble_brms_priors(): Assemble brms priors from app-level inputs
link_scale_multiplier(): Scale multiplier for the link function
priors_to_brms(): Convert prior list to brms prior object
run_scd_analysis(): Main analysis function
run_model_comparison(): Model comparison via LOO-CV
run_sequential_update(): Sequential Bayesian updating
run_sensitivity_analysis(): Run sensitivity analysis with different priors and outlier handling

Convergence diagnostics and variance-component bootstrap

compute_convergence_diagnostics(): Convergence diagnostics for Bayesian models
bootstrap_variance_components(): Parametric bootstrap for ICC CIs
prior_predictive_check(): Prior predictive check for a fitted SCD model

Posterior predictive checks on visual-analysis features and overlay draws

posterior_predictive_features(): Posterior predictive checks on single-case visual-analysis features
ppc_overlay_draws(): Posterior-predictive line draws for a spaghetti overlay

Family-native posterior decision summaries, phase contrasts, and meta-analytic summary

summarize_fixed_effect_decisions(): Posterior decision summary for fixed-effect coefficients
summarize_per_case_effects(): Per-case posterior decision summary
summarize_cross_case_consistency(): Posterior probability of consistent intervention direction across cases
summarize_variance_components(): Posterior summary of variance components
compute_phase_contrast(): Posterior phase contrast on the response scale
summarize_phase_descriptives(): Pre-model descriptive phase summary
calculate_meta_summary(): Calculate meta-analytic summary statistics

Simulation-based power for SCD multilevel Bayesian models

De-identified process-log events, sequences, and transition summaries

plog_event(): Construct a de-identified process-log event
plog_write(): Append a de-identified event to the store (no-op unless enabled)
plog_read(): Read the event store into a flat data frame of action rows
plog_summary(): Per-session and per-action event counts
plog_sequences(): Per-session ordered action sequences
plog_seq_matrix(): Right-padded action matrix for TraMineR sequence objects
plog_transition_matrix(): First-order Markov transition probabilities between consecutive actions
plog_time_to(): Time from a session's first event to a target action (right-censored)
plog_inter_event(): Within-session inter-event gaps in seconds

Gemini methodology and prior-elicitation dialog turns

gemini_ask(): Context-aware Gemini methodology assistant turn
gemini_dialog_turn(): Prior co-design (elicitation) Gemini dialog turn
sanitize_llm_text(): Strip control characters and cap length of user-supplied LLM text

scdbayes scdbayes-package: scdbayes: Bayesian Single-Case Design Analysis
print(<scd_analysis>): Print method for scd_analysis objects
.match_var(): Match a column name (case-insensitive) against a pattern list