This article was accepted into the corpus but its outbound wikilinks were never NER-processed — typical at the deepest BFS hop or when the run's entity cap was reached. No expansion funnel to show.
| pPXF | |
|---|---|
| Name | pPXF |
| Author | Michele Cappellari |
| Released | 2004 |
| Programming language | Python, originally IDL |
| License | MIT |
pPXF is a widely used software tool for extracting stellar kinematics and stellar population information from integrated-light spectra of galaxies, star clusters, and stellar populations. It fits observed spectra by convolving template spectra with parametric line-of-sight velocity distributions and by modeling additive and multiplicative corrections, enabling measurements of velocity, velocity dispersion, higher-order Gauss–Hermite moments, and population parameters. The method underpins analyses in observational programs and surveys across many facilities and collaborations.
pPXF was developed to provide robust, flexible spectral fitting for data from instruments and projects such as the Hubble Space Telescope, Very Large Telescope, Keck Observatory, Sloan Digital Sky Survey, and integral-field units like SAURON, MUSE, and VLT FLAMES. It is applied in studies of objects ranging from the Milky Way bulge and Andromeda to distant galaxies in surveys like CALIFA, MaNGA, and ATLAS3D. The software interacts with stellar libraries and synthesis models including MILES, E-MILES, STELIB, BaSeL, BC03, and Padova isochrones to reconstruct populations and kinematics. Teams affiliated with institutions such as the European Southern Observatory, Max Planck Institute for Astronomy, Institute of Astronomy, Cambridge, Space Telescope Science Institute, and Harvard-Smithsonian Center for Astrophysics have integrated pPXF into pipelines.
The core algorithm fits an observed spectrum by convolving a linear combination of template spectra with a parameterized line-of-sight velocity distribution based on Gauss–Hermite series, allowing recovery of velocity, dispersion, and asymmetric or kurtotic deviations. The approach builds on techniques used in studies associated with researchers from Padova Observatory, University of Oxford, University of Cambridge, Princeton University, and University of California, Santa Cruz. Regularization schemes borrow concepts similar to those in inverse problems tackled by groups at Caltech, Institute for Advanced Study, and ETH Zurich. The method uses non-linear optimization strategies akin to those employed in analyses by teams at NASA, CNRS, Leiden Observatory, and University of Tokyo to marginalize over continuum terms and instrumental line-spread functions. pPXF supports additive polynomials, multiplicative polynomials, and template masking, techniques also used in work from Carnegie Institution for Science, Observatoire de Paris, and University of Toronto.
Originally written in IDL, the package was ported and expanded in Python with dependencies and interfaces that mesh with libraries and tools from projects such as NumPy, SciPy, Astropy, Matplotlib, and scikit-learn. The codebase and distribution practices echo those of other astronomy software maintained at GitHub by collaborations including Astropy Project, yt Project, and SunPy. pPXF’s ecosystem integrates with spectral synthesis platforms from SYN++, STARLIGHT, STECKMAP, and FIREFLY, and is used alongside analysis tools developed at European Southern Observatory pipelines and survey teams like SDSS Collaboration. The software has been cited and adapted in research by groups at University of California, Berkeley, Columbia University, University of Oxford, University of Edinburgh, and Imperial College London.
Researchers use pPXF to derive rotation curves, velocity dispersion profiles, and higher-order kinematic maps for systems such as NGC 253, M87, NGC 1277, and M31; to measure supermassive black hole masses informed by studies at institutions like Max Planck Institute for Astrophysics, Harvard University, and University of Texas at Austin; and to trace stellar population ages and metallicities in works connected to Royal Observatory Edinburgh, Observatoire de Lyon, and Instituto de Astrofísica de Canarias. The tool supports investigations into galaxy formation and evolution topics studied in large programs such as COSMOS, CANDELS, and 3D-HST, and is used in synergy with dynamical modeling approaches from groups at University of Michigan, Rutgers University, and University of Leiden. It also assists analyses of resolved-star spectroscopy in surveys like APOGEE, RAVE, and GALAH.
Validation studies compare pPXF outputs against independent methods and instruments, drawing on benchmarks from collaborations like ATLAS3D, SAMI, CALIFA, and MaNGA. Performance assessments reference simulated and real datasets similar to those produced by Kepler, GALEX, Spitzer Space Telescope, and Chandra X-ray Observatory campaigns to test sensitivity to signal-to-noise, spectral resolution, and template mismatch. Cross-validation with full-spectrum fitting codes from teams at Max Planck Institute for Extraterrestrial Physics, University of Groningen, and University of Porto demonstrates consistency for velocity and dispersion recovery when using libraries such as MILES and E-MILES. Optimization for large surveys benefits from computational practices established at Lawrence Berkeley National Laboratory, Argonne National Laboratory, and CERN-affiliated computing centers.
Limitations include sensitivity to template mismatch when using incomplete libraries like early versions of STELIB or BaSeL, degeneracies in age–metallicity recovery noted in studies from Padova Observatory and Instituto de Astrofísica de Canarias, and challenges in low signal regimes encountered in deep-field surveys like UltraVISTA and Hubble Ultra-Deep Field. Extensions and enhancements have introduced regularized population recovery, emission-line fitting modules, and non-parametric line-of-sight velocity distributions developed in collaborations involving University of Milan, University of Bologna, National Astronomical Observatory of Japan, and Max Planck Institute for Astronomy. Ongoing integrations connect pPXF-style fitting to machine-learning frameworks used by groups at Google Research, Facebook AI Research, DeepMind, and academic labs at MIT and Stanford University to accelerate inference for upcoming facilities such as James Webb Space Telescope and next-generation extremely large telescopes.
Category:Astronomical software