Model GUI

PyDDM additionally features a GUI which allows you to manually tweak the parameters and instantaneously see the change in the model. This is useful for learning about the DDM, for gaining an intuition on new models, and for testing new models before fitting them.

The radio buttons on the left may be used to show the conditional fit of the model on only specific task parameters. The sliders on the right may be used to control the parameter fit values. The function to plot may also be customized; see plot.model_gui() for more details.

There are two versions, one for Jupyter notebooks (plot.model_gui_jupyter()) and one for a native interface (plot.model_gui()). Both have the same features and can be used interchangably based on your preference.

Try it out on Google Colab!

What can I do if the model GUI doesn’t work?

Additional views

The default view for the model GUI is plot.plot_fit_diagnostics(), which shows the response time distribution. However, other options are available, including:

Psychometric function

The psychometric function can be viewed by generating a view function with the function plot.plot_psychometric(). The psychometric is constructed with some condition on the x axis, and the probability of a correct response on the y axis. For instance, for the condition “coh”:

pyddm.plot.model_gui_jupyter(m, sample=samp, plot=pyddm.plot.plot_psychometric('coh'))

You can also split this by some other condition, for instance, a high or low attention condition named “attn”:

pyddm.plot.model_gui_jupyter(m, sample=samp, plot=pyddm.plot.plot_psychometric('coh', split_by_condition='attn'))

In addition, this GUI is also compatible with Jupyter notebooks.

Chronometric function

The chronometric function can be viewed by generating a view function with the function plot.plot_chronometric(). It operates the same way as the psychometric function.

Bound shape

For complicated forms of collapsing bounds, it can be useful to view the shape of the bound as it changes with different parameters. This can be accomplished in the model GUI using:

pyddm.plot.model_gui_jupyter(m, plot=pyddm.plot.plot_bound_shape)