PyDDM Cookbook

Below are recipies for how to get specific models and features to work in PyDDM. If you have not yet worked through the Quickstart guide, please do that first! This will give you the fundamentals for understanding the models below, demonstrated using leaky integration, collapsing bounds, and a coherence-dependent drift rate.

Common models

• Attention DDM (ADDM)

• Reinforcement learning DDM (RL-DDM)

• DDM with parameter variability (“full DDM”)

• Weibull collapsing bounds

• Multi-sensory integration

Model components

• Biased drift rate

• Changing drift rate

• Non-linear drift rate

• Unique moment-to-moment drift rate on each trial (e.g., matching EEG)

• Biased starting position

• Starting position variability

• Non-decision time variability

• Urgency signal as a multiplicative gain

• Attractor states

How-to

• Parallelization

• Stimulus coding vs accuracy coding

• Change the fitting algorithm

• Visualize the evolution of the PDF over time

• Change the objective function used for fitting

Object-oriented interface

The object-oriented interface is no longer recommended. It is not (and never will be) deprecated, but it is more complicated to use than the gddm() function.

• Quickstart guide for the object-oriented interface

• Cookbook for the object-oriented interface