FAQ

Which acquisition systems does pyfli support? ICCD, gated-SPAD (SwissSPAD2/SwissSPAD3), TCSPC, and more are natively supported. DataOperations auto-detects the format from the input file. If a format you need isn’t listed, feel free to open a discussion — the developers check in periodically and add support where possible.

Do I need a GPU? No. All analytical fitting methods (NLSF, MLE, RLD, Laguerre, phasor) run on CPU. A GPU is only used by the optional deep-learning inference backends — install the extra with pip install "pyfli-lib[gpu]" if you need it.

Why is the package pyfli-lib but I import pyfli? The name pyfli was already taken on PyPI, so the distribution is published as pyfli-lib while the importable module keeps its natural name, pyfli.

Which fitting method should I use? It depends on your data. NLSF is the standard general-purpose choice; MLE is better suited to low-photon-count data; RLD is the fastest option for real-time or high-frame-rate processing. Phasor analysis is model-free and useful for quick visual species separation without fitting a model at all.

How do I cite pyfli? See the Citation page for the full BibTeX entry, or cite the repository directly: vkp217/pyfli-pkg.

Where do I report a bug or request a feature? On the GitHub issue tracker.

Where can I get help? Open an issue on GitHub, or email support@pyfli.org or pyfli4lifetime@gmail.com.