This codebase contains all of the stimuli generating procedures using in: "A peak-end rule characterizes subjective attentional effort during moment-by-moment visual processing"

• Clone this repository
• Run julia --projec=. from inside the repo directory
• Run using Pkg; Pkg.instantiate()

Each experiment has it's own script or folder, starting with expX_*. Output will be generated under output in the root directory of the repo. Each experiment script can be run by simply invoking julia <SCRIPT>. For experiments 3 and 4, you can run julia <FOLDER>/trials.jl.

The other files either implement common functionality used in several experiments or provide debugging tools to help design new experiments. For example render_dataset.jl can be modified to render a dataset to images.

Julia manages package dependencies with Project.toml and Manifest.toml

In the /output folder, each run of an experiment script will generate the following set of files:

• script.jl: a backup of the script invoked to create the dataset
• dataset.json: The animation data for the set of trials (format described below)
• examples.json: Animation data for the example trials.
• trial_list.json : (Optional), The indices mapping animation data to trial IDs.
• metrics.json : The time course for each motion statistic used to implement time-varying difficulty.

The dataset.json file consists of a list of trials. Each trial contains an entry of positions, which contains list of time points. Each time point is a list of 2D points (one for each object).

The examples.json file contains the same format, but with fewer animations (usually only 3-4).