simtrial-fundamentalslisted

# simtrial Fundamentals ## When to Use This Skill - Simulating time-to-event (survival) clinical trial data - Generating piecewise exponential failure/dropout times - Modeling delayed treatment effects or non-proportional hazards - Performing weighted logrank tests (Fleming-Harrington, Magirr-Burman) - Running MaxCombo tests for non-proportional hazards - Simulating group sequential designs - Calculating RMST or milestone endpoints ## Package Overview **simtrial** by Merck provides fast, extensible clinical trial simulation for time-to-event endpoints. Key features: - Piecewise exponential distributions for flexible hazard modeling - Built-in support for non-proportional hazards scenarios - Integration with gsDesign2 for group sequential designs - Parallel computation via doFuture/foreach - Pipe-friendly API using data.table for performance ## Core Data Generation Functions ### sim_pw_surv() - Main Simulation Function Generates stratified time-to-event outcome randomized trial data. ```r sim_pw_surv( n = 100, # Total sample size stratum = data.frame( # Stratum definitions stratum = "All", p = 1 # Prevalence/probability ), block = c(rep("control", 2), rep("experimental", 2)), # Randomization block enroll_rate = data.frame( # Enrollment rates by period rate = 9, duration = 1 ), fail_rate = data.frame( # Failure rates by stratum/treatment/period stratum = rep("All", 4), period = rep