simtrial

Clinical Trial Simulation

Guide users through clinical trial simulation using the simtrial R package. Use this skill when the user asks about: simulating survival trials, simfix, sim_pw_surv, cutting data at calendar or event times, weighted logrank tests, MaxCombo tests, or simulation-based power.

Clinical Trial Simulation with simtrial

Note: This skill targets simtrial >= 1.0.2 (main branch at github.com/Merck/simtrial).

API reference

Vendored function docs: references/llms.txt
Full function docs (local v1.0.2): references/llms_local.txt
Workflow patterns: references/code_patterns.md

Key functions

Simulation

sim_pw_surv() - Simulate piecewise exponential survival data (individual patient data)
sim_fixed_n() - Fixed-sample simulation with analysis pipeline (simpler, less flexible)
sim_gs_n() - Group sequential simulation (integrates with gsDesign2 designs)
to_sim_pw_surv() - Convert simple rate format (control rate + HR) to sim_pw_surv format

Data cutting

cut_data_by_date() - Cut simulated data at a calendar date
cut_data_by_event() - Cut simulated data at a target event count
get_cut_date_by_event() - Find calendar date for a target event count
get_analysis_date() - Get analysis date from multiple criteria (events, calendar time, follow-up)
create_cut() - Create a cutting function for use in sim_gs_n pipelines

Statistical tests

wlr() - Weighted logrank test (single weight function)
maxcombo() - MaxCombo test (multiple FH weight functions, correlation-adjusted p-value)
rmst() / rmst_two_arm() / rmst_single_arm() - Restricted mean survival time
milestone() - Milestone analysis (survival difference at fixed time)
multitest() - Apply multiple tests to one dataset
create_test() - Create a parameterized test function for use in pipelines

Weight functions

fh() - Fleming-Harrington weights (rho, gamma)
mb() - Magirr-Burman weights (delay period for NPH)
early_zero() - Early zero weight function
wlr_weight() - General WLR weight specification

Utilities

counting_process() - At-risk/event tables from survival data (for custom analyses)
rpwexp() - Random piecewise exponential generation
rpwexp_enroll() - Random piecewise enrollment times
fit_pwexp() - Fit piecewise exponential model
randomize_by_fixed_block() - Block randomization

Output

summary() - Summarize sim_gs_n results (power, events, timing)
as_gt() - Convert summary to gt table

Example datasets

ex1_delayed_effect through ex6_crossing - Pre-built NPH scenarios from Cross-Pharma Working Group

Workflow patterns

For detailed code templates, read references/code_patterns.md.

Topics covered: - Generating survival data with sim_pw_surv() (PH and NPH) - Data cutting (calendar, event, flexible criteria) - Weighted logrank tests (wlr()) and MaxCombo (maxcombo()) - Milestone and RMST tests - Multiple tests with multitest() and create_test() - Fixed-sample simulation with sim_fixed_n() (timing_type options) - Group sequential simulation with sim_gs_n() (multiple tests, boundary updating) - Integration with gsDesign2 designs (events from design, original_design parameter) - Weight functions (Fleming-Harrington, Magirr-Burman, early zero) - Rate format conversion with to_sim_pw_surv() - Flexible analysis timing with get_analysis_date() / create_cut() - Stratified simulations - Standalone wlr() with illness-death model ADTTE data - Example NPH datasets

Important design considerations

maxcombo must be used alone in sim_gs_n(): it cannot be combined with other tests in the same test list
to_sim_pw_surv(): Use this to convert the simpler rate format (control rate + HR) to the treatment-specific format needed by sim_pw_surv()
sim_gs_n() + original_design: Pass a gsDesign2 design object to get boundaries updated based on actual vs planned information fraction
ia_alpha_spending: Controls how alpha is spent when observed events differ from planned (“min_planned_actual” is conservative default)
fa_alpha_spending = "full_alpha": Spends full alpha at final analysis (default); use “info_frac” for event underrunning scenarios
create_cut() and create_test(): These factory functions are essential for building sim_gs_n() pipelines
Example datasets: Columns are id, month, evntd, trt — need renaming to tte, event, treatment for use with wlr() / maxcombo()
Standalone wlr(): Can be used outside sim_gs_n() with any data having tte, event, stratum, treatment columns. Returns positive Z when experimental is better. See references/code_patterns.md for illness-death model integration.

Code Patterns

Code Patterns for simtrial

Note: These patterns target simtrial >= 1.0.2 (main branch at github.com/Merck/simtrial).

Generating survival data with sim_pw_surv
Data cutting
Weighted logrank test (wlr)
MaxCombo test
Milestone and RMST tests
Multiple tests with multitest
Fixed-sample simulation with sim_fixed_n
Group sequential simulation with sim_gs_n
Integrating with gsDesign2 designs
Weight functions
Converting rate formats with to_sim_pw_surv
Flexible analysis timing with get_analysis_date
Example NPH datasets
Stratified simulations

Generating survival data with sim_pw_surv

sim_pw_surv() generates individual patient data with piecewise exponential enrollment, failure, and dropout.

Basic usage (proportional hazards)

library(simtrial)

# Simulate 400 patients, 1:1 randomization
data <- sim_pw_surv(
  n = 400,
  stratum = data.frame(stratum = "All", p = 1),
  block = c(rep("control", 2), rep("experimental", 2)),
  enroll_rate = data.frame(rate = 25, duration = 16),
  fail_rate = data.frame(
    stratum = rep("All", 2),
    period = c(1, 1),
    treatment = c("control", "experimental"),
    duration = c(100, 100),
    rate = c(log(2) / 12, log(2) / 12 * 0.7)
  ),
  dropout_rate = data.frame(
    stratum = rep("All", 2),
    period = c(1, 1),
    treatment = c("control", "experimental"),
    duration = c(100, 100),
    rate = c(0.001, 0.001)
  )
)
# Returns: stratum, enroll_time, treatment, fail_time, dropout_time, cte, fail

Delayed treatment effect (non-proportional hazards)

data <- sim_pw_surv(
  n = 400,
  stratum = data.frame(stratum = "All", p = 1),
  block = c(rep("control", 2), rep("experimental", 2)),
  enroll_rate = data.frame(rate = c(10, 30), duration = c(4, 12)),
  fail_rate = data.frame(
    stratum = rep("All", 4),
    period = rep(1:2, 2),
    treatment = c(rep("control", 2), rep("experimental", 2)),
    duration = rep(c(3, 100), 2),
    rate = log(2) / c(9, 9, 9, 18)   # HR = 1 first 3 months, then 0.5
  ),
  dropout_rate = data.frame(
    stratum = rep("All", 2),
    period = c(1, 1),
    treatment = c("control", "experimental"),
    duration = c(100, 100),
    rate = c(0.001, 0.001)
  )
)

Using to_sim_pw_surv for simpler rate specification

# Define rates in the simpler format (like sim_fixed_n)
fail_rate <- data.frame(
  stratum = "All",
  duration = c(3, 100),
  fail_rate = log(2) / c(9, 18),
  hr = c(0.9, 0.6),
  dropout_rate = 0.001
)

# Convert to sim_pw_surv format
rates <- to_sim_pw_surv(fail_rate)

data <- sim_pw_surv(
  n = 400,
  enroll_rate = data.frame(rate = c(10, 30), duration = c(4, 12)),
  fail_rate = rates$fail_rate,
  dropout_rate = rates$dropout_rate
)

Data cutting

After generating data with sim_pw_surv(), cut it at a specific calendar time or event count to create an analysis dataset.

Cut by calendar date

# Cut at 36 months from study start
analysis_data <- cut_data_by_date(data, cut_date = 36)
# Returns: tte, event, stratum, treatment (class "tte_data")

Cut by event count

# Cut when 250 events observed
analysis_data <- cut_data_by_event(data, event = 250)

Using create_cut for pipelines

create_cut() wraps get_analysis_date() arguments into a reusable function.

# Cut at planned calendar time
cut1 <- create_cut(planned_calendar_time = 24)

# Cut at target events
cut2 <- create_cut(target_event_overall = 250)

# Cut at target events but no more than 6 months extension
cut3 <- create_cut(
  target_event_overall = 300,
  max_extension_for_target_event = 6
)

# Apply cutting function to simulated data
analysis_data <- cut1(data)

Weighted logrank test (wlr)

wlr() performs a single weighted logrank test on cut data.

# Standard logrank
result <- wlr(
  data = cut_data_by_event(data, 250),
  weight = fh(rho = 0, gamma = 0)
)
result$z  # Z-statistic (negative favors experimental)

# Fleming-Harrington (0, 0.5) - emphasizes late differences
result_fh <- wlr(
  data = cut_data_by_event(data, 250),
  weight = fh(rho = 0, gamma = 0.5)
)

# With variance estimate
result_var <- wlr(
  data = cut_data_by_event(data, 250),
  weight = fh(rho = 0, gamma = 0),
  return_variance = TRUE
)

MaxCombo test

maxcombo() combines multiple Fleming-Harrington weighted logrank tests, using the maximum test statistic with a correlation-adjusted p-value.

# MaxCombo: logrank + FH(0, 0.5)
result <- maxcombo(
  data = cut_data_by_event(data, 250),
  rho = c(0, 0),
  gamma = c(0, 0.5)
)
result$z        # Z-statistics for each test
result$p_value  # p-value (accounts for correlation)

# Three-test MaxCombo
result3 <- maxcombo(
  data = cut_data_by_event(data, 250),
  rho = c(0, 0, 1),
  gamma = c(0, 1, 1),
  return_corr = TRUE   # Also return correlation matrix
)

Milestone and RMST tests

Milestone (survival difference at fixed time)

result <- milestone(
  data = cut_data_by_event(data, 250),
  ms_time = 12,            # Compare survival at 12 months
  test_type = "naive"      # Or "log-log"
)
result$estimate  # Survival difference (experimental - control)
result$z         # Z-statistic

RMST (restricted mean survival time difference)

result <- rmst(
  data = cut_data_by_event(data, 250),
  tau = 24                 # RMST up to 24 months
)

Multiple tests with multitest

Apply several tests to the same cut dataset in one call.

# Create parameterized test functions
logrank <- create_test(wlr, weight = fh(rho = 0, gamma = 0))
fh05 <- create_test(wlr, weight = fh(rho = 0, gamma = 0.5))
rmst_test <- create_test(rmst, tau = 24)
combo <- create_test(maxcombo, rho = c(0, 0), gamma = c(0, 0.5))

# Apply all tests to one dataset
cut <- cut_data_by_event(data, 250)
results <- multitest(
  data = cut,
  logrank = logrank,
  fh05 = fh05,
  rmst = rmst_test,
  combo = combo
)
# results is a named list of test outputs

Fixed-sample simulation with sim_fixed_n

sim_fixed_n() is a self-contained pipeline: simulate data, cut, and test in one call. Simpler but less flexible than sim_gs_n().

# Basic simulation with Fleming-Harrington tests
results <- sim_fixed_n(
  n_sim = 1000,
  sample_size = 400,
  target_event = 300,
  enroll_rate = data.frame(duration = c(4, 12), rate = c(10, 30)),
  fail_rate = data.frame(
    stratum = "All",
    duration = c(3, 100),
    fail_rate = log(2) / c(9, 18),
    hr = c(0.9, 0.6),
    dropout_rate = 0.001
  ),
  total_duration = 36,
  timing_type = 1:5,
  rho_gamma = data.frame(rho = c(0, 0), gamma = c(0, 0.5))
)

# timing_type controls the cutoff method:
# 1: Planned study duration (total_duration)
# 2: Time target_event is achieved
# 3: Minimum follow-up after enrollment complete
# 4: max(planned duration, target event time)
# 5: max(target event time, minimum follow-up)

Estimating power from sim_fixed_n

# Power = proportion of simulations with z < -1.96
power <- results |>
  dplyr::filter(cut == "Planned duration" & rho == 0 & gamma == 0) |>
  dplyr::summarize(power = mean(z <= qnorm(0.025)))

Group sequential simulation with sim_gs_n

sim_gs_n() simulates a group sequential trial: generate data, cut at multiple analysis times, and apply tests.

Basic group sequential simulation

library(simtrial)
library(gsDesign2)
library(gsDesign)

# Define cutting functions for 3 analyses
ia1_cut <- create_cut(target_event_overall = 150)
ia2_cut <- create_cut(target_event_overall = 225)
fa_cut <- create_cut(target_event_overall = 300)

# Simulate with logrank test
results <- sim_gs_n(
  n_sim = 1000,
  sample_size = 400,
  enroll_rate = data.frame(duration = c(4, 12), rate = c(10, 30)),
  fail_rate = data.frame(
    stratum = "All",
    duration = c(3, 100),
    fail_rate = log(2) / c(9, 18),
    hr = c(0.9, 0.6),
    dropout_rate = 0.001
  ),
  test = wlr,
  cut = list(ia1 = ia1_cut, ia2 = ia2_cut, fa = fa_cut),
  weight = fh(rho = 0, gamma = 0)
)

MaxCombo in sim_gs_n

MaxCombo must be the only test (cannot be combined with other tests in the same call).

results_combo <- sim_gs_n(
  n_sim = 1000,
  sample_size = 400,
  enroll_rate = data.frame(duration = c(4, 12), rate = c(10, 30)),
  fail_rate = data.frame(
    stratum = "All",
    duration = c(3, 100),
    fail_rate = log(2) / c(9, 18),
    hr = c(0.9, 0.6),
    dropout_rate = 0.001
  ),
  test = maxcombo,
  cut = list(ia1 = ia1_cut, ia2 = ia2_cut, fa = fa_cut),
  rho = c(0, 0),
  gamma = c(0, 0.5)
)

Multiple WLR tests in sim_gs_n

results_multi <- sim_gs_n(
  n_sim = 1000,
  sample_size = 400,
  enroll_rate = data.frame(duration = c(4, 12), rate = c(10, 30)),
  fail_rate = data.frame(
    stratum = "All",
    duration = c(3, 100),
    fail_rate = log(2) / c(9, 18),
    hr = c(0.9, 0.6),
    dropout_rate = 0.001
  ),
  test = list(
    logrank = create_test(wlr, weight = fh(rho = 0, gamma = 0)),
    fh05 = create_test(wlr, weight = fh(rho = 0, gamma = 0.5))
  ),
  cut = list(ia1 = ia1_cut, ia2 = ia2_cut, fa = fa_cut)
)

Integrating with gsDesign2 designs

Use a gsDesign2 design to drive sample size, event targets, and boundary updates in sim_gs_n().

library(simtrial)
library(gsDesign2)
library(gsDesign)

# Enrollment and failure rate assumptions
enroll_rate <- define_enroll_rate(
  duration = c(4, 12),
  rate = c(10, 30)
)

fail_rate <- define_fail_rate(
  duration = c(3, 100),
  fail_rate = log(2) / 9,
  hr = c(0.9, 0.6),
  dropout_rate = 0.001
)

# Group sequential design
design <- gs_design_ahr(
  enroll_rate = enroll_rate,
  fail_rate = fail_rate,
  alpha = 0.025,
  beta = 0.1,
  analysis_time = c(12, 24, 36),
  info_scale = "h0_info",
  upper = gs_spending_bound,
  upar = list(sf = sfLDOF, total_spend = 0.025),
  lower = gs_b,
  lpar = rep(-Inf, 3)
)

# Create cuts from the design's planned events
cuts <- list(
  ia1 = create_cut(target_event_overall = ceiling(design$analysis$event[1])),
  ia2 = create_cut(target_event_overall = ceiling(design$analysis$event[2])),
  fa = create_cut(target_event_overall = ceiling(design$analysis$event[3]))
)

# Simulate with boundary updating from original design
results <- sim_gs_n(
  n_sim = 1000,
  sample_size = ceiling(max(design$analysis$n)),
  enroll_rate = data.frame(
    duration = design$enroll_rate$duration,
    rate = design$enroll_rate$rate
  ),
  fail_rate = data.frame(
    stratum = "All",
    duration = fail_rate$duration,
    fail_rate = fail_rate$fail_rate,
    hr = fail_rate$hr,
    dropout_rate = fail_rate$dropout_rate
  ),
  test = wlr,
  cut = cuts,
  original_design = design,
  weight = fh(rho = 0, gamma = 0)
)

# Summarize results
results |> summary() |> as_gt()

Alpha spending options

# How interim alpha is spent when observed events differ from planned
results <- sim_gs_n(
  ...,
  original_design = design,
  ia_alpha_spending = "min_planned_actual",  # Default: min of planned and actual info fraction
  fa_alpha_spending = "full_alpha"           # Default: spend full alpha at final analysis
)

# Alternative: spend by actual info fraction at FA (for underrunning)
results <- sim_gs_n(
  ...,
  original_design = design,
  ia_alpha_spending = "actual",
  fa_alpha_spending = "info_frac"
)

Weight functions

Fleming-Harrington family

fh(rho = 0, gamma = 0)    # Standard logrank
fh(rho = 0, gamma = 0.5)  # Emphasizes late differences
fh(rho = 0, gamma = 1)    # Stronger late emphasis (Peto-logrank-like)
fh(rho = 1, gamma = 0)    # Emphasizes early differences (Prentice-Wilcoxon)
fh(rho = 1, gamma = 1)    # Moderate weight all around

Magirr-Burman weights (for delayed effects)

mb(delay = 4)              # Down-weight first 4 months
mb(delay = 4, w_max = 2)  # Cap maximum weight at 2
mb(delay = Inf, w_max = 2) # Magirr (2021) recommendation

Early zero weight

early_zero(early_period = 4)  # Zero weight before 4 months

Converting rate formats with to_sim_pw_surv

sim_fixed_n() and sim_gs_n() use a simpler rate format (control rate + HR). sim_pw_surv() uses treatment-specific rates. to_sim_pw_surv() converts between them.

# Simple format (control rate + HR)
fail_rate_simple <- data.frame(
  stratum = "All",
  duration = c(3, 100),
  fail_rate = log(2) / c(9, 18),
  hr = c(0.9, 0.6),
  dropout_rate = 0.001
)

# Convert to sim_pw_surv format
rates <- to_sim_pw_surv(fail_rate_simple)
rates$fail_rate    # Treatment-specific failure rates
rates$dropout_rate # Treatment-specific dropout rates

# Use in sim_pw_surv
data <- sim_pw_surv(
  n = 400,
  enroll_rate = data.frame(rate = 25, duration = 16),
  fail_rate = rates$fail_rate,
  dropout_rate = rates$dropout_rate
)

Flexible analysis timing with get_analysis_date

get_analysis_date() determines analysis timing from multiple criteria. create_cut() wraps it for use in pipelines.

Calendar-time-driven analysis

cut <- create_cut(planned_calendar_time = 36)

Event-driven analysis

cut <- create_cut(target_event_overall = 300)

Event-driven with maximum extension

# Target 300 events, but extend at most 6 months beyond planned time
cut <- create_cut(
  target_event_overall = 300,
  max_extension_for_target_event = 42  # Absolute max calendar time
)

Minimum follow-up

# At least 12 months follow-up after last enrollment
cut <- create_cut(min_followup = 12)

Combining criteria

# Analysis at the later of target events or minimum follow-up
cut <- create_cut(
  target_event_overall = 300,
  min_followup = 12
)

# Minimum time between analyses
cut_ia2 <- create_cut(
  target_event_overall = 225,
  min_time_after_previous_analysis = 6
)

Per-stratum event targets

cut <- create_cut(
  target_event_per_stratum = c("BM+" = 100, "BM-" = 150)
)

Example NPH datasets

simtrial includes pre-built datasets from the Cross-Pharma Non-Proportional Hazards Working Group.

# Delayed treatment effect
data(ex1_delayed_effect)
# Columns: id, month (tte), evntd (event indicator), trt (0=ctrl, 1=exp)

data(ex2_delayed_effect)  # Another delayed effect scenario
data(ex3_cure_with_ph)    # Cure model with proportional hazards
data(ex4_belly)           # "Belly-shaped" survival curves
data(ex5_widening)        # Widening survival curves
data(ex6_crossing)        # Crossing survival curves

# Use with wlr (need column renaming)
library(dplyr)
analysis_data <- ex1_delayed_effect |>
  rename(tte = month, event = evntd, treatment = trt) |>
  mutate(
    treatment = ifelse(treatment == 1, "experimental", "control"),
    stratum = "All"
  )
wlr(analysis_data, weight = fh(rho = 0, gamma = 0))

Stratified simulations

Generating stratified data

data <- sim_pw_surv(
  n = 600,
  stratum = data.frame(stratum = c("BM+", "BM-"), p = c(0.4, 0.6)),
  block = c(rep("control", 2), rep("experimental", 2)),
  enroll_rate = data.frame(rate = 30, duration = 20),
  fail_rate = data.frame(
    stratum = rep(c("BM+", "BM+", "BM-", "BM-"), each = 1),
    period = rep(1, 4),
    treatment = rep(c("control", "experimental"), 2),
    duration = rep(100, 4),
    rate = c(log(2)/12, log(2)/12 * 0.6,   # BM+: HR = 0.6
             log(2)/12, log(2)/12 * 0.85)   # BM-: HR = 0.85
  ),
  dropout_rate = data.frame(
    stratum = rep(c("BM+", "BM-"), each = 2),
    period = rep(1, 4),
    treatment = rep(c("control", "experimental"), 2),
    duration = rep(100, 4),
    rate = rep(0.001, 4)
  )
)

# Cut and test with stratified WLR
cut <- cut_data_by_event(data, event = 350)
wlr(cut, weight = fh(rho = 0, gamma = 0))

Stratified sim_fixed_n

results <- sim_fixed_n(
  n_sim = 1000,
  sample_size = 600,
  target_event = 350,
  stratum = data.frame(stratum = c("BM+", "BM-"), p = c(0.4, 0.6)),
  enroll_rate = data.frame(duration = c(4, 16), rate = c(10, 30)),
  fail_rate = data.frame(
    stratum = rep(c("BM+", "BM-"), each = 2),
    duration = rep(c(3, 100), 2),
    fail_rate = log(2) / 12,
    hr = c(0.6, 0.6, 0.85, 0.85),
    dropout_rate = 0.001
  ),
  total_duration = 36,
  timing_type = 1:2,
  rho_gamma = data.frame(rho = 0, gamma = 0)
)

Standalone wlr with illness-death model data

wlr() can be used standalone (outside sim_gs_n()) with data from the illness-death model. The key is mapping ADTTE columns to the wlr() format.

Column mapping

wlr() expects columns: tte, event, stratum, treatment

# From illness-death ADTTE data (columns: AVAL, CNSR, STRATUM, TRT)
logrank_z <- function(data) {
  d <- data.frame(
    tte = data$AVAL,
    event = 1L - data$CNSR,
    stratum = data$STRATUM,
    treatment = data$TRT
  )
  if (sum(d$event) < 2) return(NA_real_)
  wlr(d, weight = fh(rho = 0, gamma = 0))$z
}

Sign convention

wlr() returns positive Z when experimental is better (more events in control than expected under null). This matches the convention needed for sequentialPValue() and graphical testing.

Usage in simulation loops

adtte_18 <- cut_illness_death(sim_data, cut_date = 18)
z_pfs_18 <- logrank_z(adtte_18[adtte_18$PARAMCD == "PFS", ])
z_os_18 <- logrank_z(adtte_18[adtte_18$PARAMCD == "OS", ])

# Collect event counts for spending time
ev_pfs_18 <- sum(1L - adtte_18$CNSR[adtte_18$PARAMCD == "PFS"])
ev_os_18 <- sum(1L - adtte_18$CNSR[adtte_18$PARAMCD == "OS"])

Clinical Trial Simulation with simtrial

API reference

Key functions

Simulation

Data cutting

Statistical tests

Weight functions

Utilities

Output

Example datasets

Workflow patterns

Important design considerations

Code Patterns

Code Patterns for simtrial

Table of Contents

Generating survival data with sim_pw_surv

Basic usage (proportional hazards)

Delayed treatment effect (non-proportional hazards)

Using to_sim_pw_surv for simpler rate specification

Data cutting

Cut by calendar date

Cut by event count

Using create_cut for pipelines

Weighted logrank test (wlr)

MaxCombo test

Milestone and RMST tests

Milestone (survival difference at fixed time)

RMST (restricted mean survival time difference)

Multiple tests with multitest

Fixed-sample simulation with sim_fixed_n

Estimating power from sim_fixed_n

Group sequential simulation with sim_gs_n

Basic group sequential simulation

MaxCombo in sim_gs_n

Multiple WLR tests in sim_gs_n

Integrating with gsDesign2 designs

Alpha spending options

Weight functions

Fleming-Harrington family

Magirr-Burman weights (for delayed effects)

Early zero weight

Converting rate formats with to_sim_pw_surv

Flexible analysis timing with get_analysis_date

Calendar-time-driven analysis

Event-driven analysis

Event-driven with maximum extension

Minimum follow-up

Combining criteria

Per-stratum event targets

Example NPH datasets

Stratified simulations

Generating stratified data

Stratified sim_fixed_n

Standalone wlr with illness-death model data

Column mapping

Sign convention

Usage in simulation loops