Cite the software, preserve the analysis¶

The published eyetrajectoriespy 1.2.0 package is software for continuous and functional eye-tracking trajectories. A software citation is not evidence that a particular scientific estimator is qualified for a new research design. Cite the software version you actually used and, separately, the statistical method or validation evidence relevant to your analysis.

Published 1.2.0 release Public API and version boundaries Research reporting checklist

Software citation¶

The following fields come from the repository's CITATION.cff and published release:

Balaskas, S. (2026). eyetrajectoriespy: Functional and Continuous Trajectory Analysis for Eye-Tracking Data in Python (Version 1.2.0) [Computer software]. https://github.com/stefanosbalaskas/eyetrajectoriespy

For a BibTeX-compatible software entry, use:

@software{balaskas_eyetrajectoriespy_2026,
  author  = {Balaskas, Stefanos},
  title   = {eyetrajectoriespy: Functional and Continuous Trajectory Analysis for Eye-Tracking Data in Python},
  year    = {2026},
  version = {1.2.0},
  url     = {https://github.com/stefanosbalaskas/eyetrajectoriespy}
}

The release's CFF metadata currently does not declare a DOI. Do not manufacture one. If your software environment differs from the published 1.2.0 distribution (e.g. an unreleased development commit or repository-only adapter example), cite the exact commit and explicitly name it as development software.

Which evidence should accompany a citation?¶

What you use	Additional source to report	What not to claim
Published 1.2.0 stable API	Release notes, actual public function and its documented method	That the package version alone validates hardware or study design
One of ten module-scoped W1–W4 routes	Workflow contracts and report bundle	That the orchestration wrapper automatically validates all estimates
Sparse/irregular or coordinate-specific observations	Native irregular guide and sparse FPCA	That gaps were observed or imputed without declaring a method
Functional confidence bands	Inference assumptions and validation ledger	That W3 whole-function bands are generally calibrated
Repository SyeLink, pymovements or BIDS examples	Integration status, exact upstream version and commit	That these previews are released, qualified vendor/device adapters
Experimental E1–E4 helpers	Experimental development boundary	That an unreleased helper is stable or publishable inference

Minimum reproducibility record¶

Keep the following together with any figure or research table:

Source and consent: acquisition system, original-export provenance, eligibility, and whether the files are recorded or simulated.
Software provenance: Python version, exact eyetrajectoriespy distribution version or commit, optional libraries and acquisition parser versions.
Measurement contract: gaze eye, timestamp origin and unit, x/y coordinate system and screen geometry, sample masks, transformations and any gap handling.
Independent units: participants, nested trials, experiment conditions, exclusion counts and the unit used for validation or resampling.
Scientific analysis: predeclared question, estimand, method, hyperparameters, diagnostics, sensitivity checks, and inferential scope.
Outputs: executable script, random seeds, environment record, figures and a retained reproducibility bundle.

The downloadable seeded teaching example illustrates a reproducible path but contains fabricated values—not empirical eye-tracking records.

Scientific uncertainty remains conditional

Whole-function inference under the open W3 qualification issue #289 and field identification under missing-gaze issue #296 must not be described as generally validated. See limitations and failure modes.

Prefer a persistent research record¶

For replication, save a frozen repository commit or tagged release, the CFF citation, a dependency lock/environment manifest, scripts, data-permission terms and clearly labelled intermediate outputs. A GitHub repository link is the available source identifier here; no archival identifier or DOI is asserted.

Begin a first analysis · Visual methods explorer · Stable versus experimental status