Pyxel 1.0: an open source Python framework for detector and end-to-end instrument simulation

Abstract

Detector modeling is becoming more and more critical for the development of new instruments in scientific space missions and ground-based experiments. Modeling tools are often developed from scratch by each individual project and not necessarily shared for reuse by a wider community. To foster knowledge transfer, reusability, and reliability in the instrumentation community, we developed Pyxel, a framework for the simulation of scientific detectors and instruments. Pyxel is an open-source and collaborative project, based on Python, developed as an easy-to-use tool that can host and pipeline any kind of detector effect model. Recently, Pyxel has achieved a new milestone: the public release and launch of version 1.0, which simplified third-party contributions and improved ease of use even further. Since its launch, Pyxel has been experiencing a growing user community and is being used to simulate a variety of detectors. We give a tour of Pyxel's version 1.0 changes and new features, including a new interface, parallel computing, and new detectors and models. We continue with an example of using Pyxel as a tool for model optimization and calibration. Finally, we describe an example of how Pyxel and its features can be used to develop a full-scale end-to-end instrument simulator.

AutoSourceID-FeatureExtractor

“…In future studies, we aim to address this discrepancy between synthetic and real images more effectively. By integrating software such as Pyxel (Arko et al 2022), we hope to train on images that more closely resemble real-world scenarios, minimising the potential performance drop when a model trained on synthetic images is used for prediction on real-world data. Ultimately, these advancements will serve to close the reality gap, enhancing the model's applicability and reliability in real-world applications (Caron et al 2023).…”

“…As the intersection of machine learning and astronomy continues to evolve, it is essential to recognise the burgeoning role of these sophisticated tools. Tools such as Pyxel (Arko et al 2022) and ScopeSim (Leschinski et al 2020) are leading this advancement, striving to produce images with an unparalleled level of detail that encapsulate everything from subtle celestial features to a broad spectrum of observational artefacts. While the astronomical landscape may not be dominated by these synthetic images, their increasing fidelity makes them invaluable assets for machine learning training.…”

PlatoSim: an end-to-end PLATO camera simulator for modelling high-precision space-based photometry

“…SIXTE; Dauser et al 2019), optical (CHEOPSim;Futyan et al 2020, SIMCADO;Leschinski et al 2016), (near)infrared (MIRISim; Klaassen et al 2021, Specsim;Lorente et al 2006), and all the way to the radio (pyuvsim; Lanman et al 2019). Also multi-purpose software packages exist, such as MAISIE (O'Brien et al 2016) and SOPHISM (Blanco Rodríguez et al 2018), and simulation frameworks, such as Pyxel (Arko et al 2022), which are specifically designed for detectors.…”