Here are some information on how to get started with the new function for the 7-parameter drift diffusion model.

The paper introducing the new function to Stan is Henrich et al. (2023). There is the theoratical background and a description on how to set up the diffusion model in Stan.

The function is available in Stan (since release 2.35.0, June 2024).

Please feel free to use the function for your purposes and report any bugs or ask questions to the authors:

Franziska Henrich (franziska.henrich@psychologie.uni-freiburg.de) and Valentin Pratz (pratz@stud.uni-heidelberg.de,

or open an issue on Github referring to the seven-parameter diffusion model: https://github.com/stan-dev/math/issues.

Remark

1) The function is now called in another way as described in the paper: To call the function use wiener() or wiener_lpdf() instead of wiener_full() or wiener_full_lpdf(), respectively (also have this in mind when you look at Valentines example below). 2) The ordering of the variables is the following: wiener_lpdf(a, t0, w, v, sv, sw, st0). In an earlier version of the code we used another ordering. But this will cause weird errors. (Again, in Valentines code below, the old ordering is present. Don't be confused of this.)

Illustration of the function's behavior

Valentin followed the three-level setup by Boehem et al. (2018) and provides a hands-on illustration of the function's behavior in three blog entries:
Level 1
Level 2
Level 3

Example R-Code for cmdstanr.

data.rds of the form: condition|resonse|reaction_time

library(cmdstanr)
library(readr)
set_cmdstan_path("path/to/cmdstan")
file <- file.path("stan/wiener_lpdf.stan")
mod <- cmdstan_model(file)
init.full = function(chains=4, N=200){
  L = list()
  for (i in 1:chains) {
    L[[i]] = list(
      a = 1,
      zr_m = 0.5,
      v_m = c(0,0),
      t0_m = 0.1,
      zr_s = 0.1,
      v_s = 0.2,
      t0_s = 0.1
    )
  }
  return (L)
}

sim.data = readRDS("path/to/data.rds")
stan.data = list(
  N = nrow(sim.data),     # Number of Trials
  cnd = sim.data$cnd,     # stimulus type (1,2)
  Ncnds = length(unique(sim.data$cnd)), # Number of conditions
  rt = sim.data$rt,       # Rt in seconds!
  resp = sim.data$resp,    # response (0=lower threshold, 1=upper threshold)
  precision = 1e-3,
  parallel = 0
)
fit <- mod$sample(
  data = stan.data,
  seed = 123,
  chains = 4,
  parallel_chains = 4,
  refresh = 10,
  iter_warmup = 200,
  iter_sampling = 300,
  init = init.full()
)

References

Boehm, U., Annis, J., Frank, M. J., Hawkins, G. E., Heathcote, A., Kellen, D., Krypotos, A.-M., Lerche, V., Logan, G. D., Palmeri, T. J., van Ravenzwaaij, D., Servant, M., Singmann, H., Starns, J. J., Voss, A., Wiecki, T. V., Matzke, D., & Wagenmakers, E.-J. (2018). Estimating across-trial variability parameters of the Diffusion Decision Model: Expert advice and recommendations. Journal of Mathematical Psychology, 87(4), 46–75. https://doi.org/10.1016/j.jmp.2018.09.004

Henrich, F., Hartmann, R., Pratz, V., Voss, A., & Klauer, K. C. (2023). The Seven-parameter Diffusion Model: An Implementation in Stan for Bayesian Analyses. Behavior Research Methods. https://doi.org/10.3758/s13428-023-02179-1