DiffusionModels

This repository provides a Julia interface to Jan Drugowitsch's C++ Diffusion model library.

No guarantee is provided for the correctness of the implementation.

The code is licensed under the New BSD License.

Usage

The function fpt returns the first passage time densities up to tmax for the upper and lower bounds of a diffusion model with time-varying drift, time-varying variance and time-varying asymmetric bounds.

rand returns random numbers from a diffusion model, either with or without a non-decision time.

Example

In this example, we use a constant drift, constant standard deviation and constant symmetric bounds (ndt currently has no effect on first-passage time).

μ = ConstDrift(1.2)
σ = ConstSigma(1)
bounds = ConstSymBounds(2,-2)
ndt = NonDecisionTime(lower = 2, upper = 3)
dm = DiffusionModel(μ, σ, bounds, ndt, Δt = 0.01)

First passage time densities

upper, lower = fpt(dm, tmax = 10.0)

Random samples

choice, rt = rand(dm, 100)