pyFlies

A Domain-Specific Language (DSL) for designing experiments in psychology

---

pyFlies is a language tailored for designing of experiments in psychology. These kind of languages are often called Domain-Specific Languages - DSLs.

The aim of the language is to be highly readable and simple to learn and to focus on capturing the essence of the experiment while leaving the details to the compiler/code generator.

A code for various run-time platforms can be generated from the experiment description. Currently PsychoPy is fully supported and we plan to build generators for other targets.

Features:

• High-level. Easy to write and read. Experiments can be defined in minutes.
• From experiment description a source code for various platforms can be automatically generated.
• Declarative language. Specify what needs to be done and leave how part to the pyFlies compiler.
• Integrates in VS Code, one of the most popular code editor today with a huge number of extensions for all sorts of file editing tasks.
• Written in Python programming language. Easy to extend. Generators are plugins which can be developed independently.
• Fully free and open source. GPL license. Hosted on github. Easy to contribute to. Contributions are welcomed!

Overview video

Click on the image bellow to see a short (~3 min) overview video. For a longer introduction please see tutorial videos.

Getting started

Installation

Install Python and check that it is available on the command line by running:

python --version

It is recommended to use Python virtual environments to isolate different set of Python libraries. Create virtual environment for pyFlies by running:

python -m venv pyflies-venv

This will create folder pyflies-venv where your libraries will be installed.

You need to activate virtual environment before usage:

source pyflies-venv/bin/activate    (for Linux and other POSIX systems)
pyflies-venv\Scripts\activate.bat    (for Windows)

Now, you can install pyFlies and generator for PsychoPy with:

pip install pyflies-psychopy

To verify that pyFlies is installed you can run:

textx list-generators

You can see in the output that the generator pyflies -> psychopy is available.

pyFlies specifications are pure text and can be edited by any textual editor but for a good experience (especially with tables) it is recommended that pyFlies extension for VS Code is used.

Install VS Code either for you OS package manager or by going to VS Code download page and downloading package for your operating system. In the list of extensions find pyFlies and click on install.

Video tutorials

The best way to start with pyFlies is by watching some of our video tutorials.

Try examples

Clone or download pyFlies repo. Unpack and load examples from examples folder in the VS Code editor. Update experiment definition to your taste. Generate and run experiment.

Planned features

Besides language improvement which we are constantly work on, there are two bigger features that will greatly improve overall user experience in working in the pyFlies editor:

• Running generators from the context menu in VS Code and getting feedback and errors.
• "Smart" code-completion and code navigation. Currently, we provide only code snippets which can help a lot but implementing Language Server Protocol for pyFlies we can get much better editing experience. For example, when you type exec -> editor can offer just the components, or when you are inside the component parameters editor can offer just the parameters for that component (without those that are already specified) with explanation extracted from the component description in the tooltip. Furthermore, navigation will enable to click on a variable and jump to its definition.

Discuss, ask questions

For all questions, ideas, feature requests and bug report please use the GitHub issue tracker.

Citing

If you are using PyFlies in your research or you are referring to PyFlies in your papers please cite this paper:

Dejanović, I.; Dejanović, M.; Vidaković, J.; Nikolić, S. PyFlies: A Domain-Specific Language for Designing Experiments in Psychology. Appl. Sci. 2021, 11, 7823. https://doi.org/10.3390/app11177823

@Article{app11177823,
AUTHOR = {Dejanović, Igor and Dejanović, Mirjana and Vidaković, Jovana and Nikolić, Siniša},
TITLE = {PyFlies: A Domain-Specific Language for Designing Experiments in Psychology},
JOURNAL = {Applied Sciences},
VOLUME = {11},
YEAR = {2021},
NUMBER = {17},
ARTICLE-NUMBER = {7823},
URL = {https://www.mdpi.com/2076-3417/11/17/7823},
ISSN = {2076-3417},
DOI = {10.3390/app11177823}
}

Screenshots (click for a popup)

Editing specification

Experiment is created and modified in a convenient to use editor.

Generated log

From an experiment specification a full log of the experiment flow is generated which can be used to investigate and debug the experiment.

Generated PsychoPy code

From an experiment specification Python code for PsychoPy is generated which is used to run the experiment.

pyFlies icon is based on an icon licensed by CC BY 3.0 from Icon Fonts.