Can the DynNom function in R create a dynamic nomogram from a Cox model with time varying coefficients

Question

Recently, I learnt about DynNom function which allows one to generate a dynamic nomogram. When I attempted to generate a dynamic nomogram from a Cox model allowing for time varying coefficient after using the survSplit function, an error reading “start/stop notation not supported”.

Below is a sample dataset and code to replicate the error.

# Load necessary libraries
  library(survival)
  library(rms)
  library(DynNom)

# Example: Simulated dataset
  set.seed(123)
  n <- 200
  data <- data.frame(
    id = 1:n,
    age = rnorm(n, 60, 10), # Age as a covariate
    sex = sample(c(0, 1), n, replace = TRUE), # Sex as a binary covariate
    start = 0,
    stop = rexp(n, rate = 0.1), # Random survival times
    event = sample(c(0, 1), n, replace = TRUE) # Random event indicator
   )

# Split dataset to incorporate time-varying covariates
  data_split <- survSplit(
    Surv(start, stop, event) ~ ., 
    data = data, 
    cut = quantile(data$stop, probs = c(0.25, 0.5, 0.75)), # Time split points
    episode = "time_period" # New variable for time periods
)

# Cox model with time-varying coefficients
  cox_model <- coxph(
    Surv(start, stop, event) ~ age + sex + time_period:age + time_period:sex,
    data = data_split
  )


# I tried the code below

# Dynamic nomogram
  dd <- datadist(data_split)
  options(datadist = "dd")

  dyn_nom <- DynNom(cox_model, data_split)

And an error note above was returned.

Any suggestions about how a dynamic nomogram can be generated with data from survSplit?

Thank you very much for your time.