Unbinned maximum-likelihood estimators for low-count data: Applications to faint X-ray spectra in the Taurus Molecular Cloud

Abstract

Traditional binned statistics such as ${\chi }^2$ suffer from information loss and arbitrariness of the binning procedure. We point out that the underlying statistical quantity (the log likelihood $L$) does not require any binning beyond the one implied by instrumental readout channels, and we propose to use it for low-count data. The performance of $L$ in the model classification and point estimation problems is explored by Monte-Carlo simulations of Chandra and XMM X-ray spectra, and is compared to the performances of the binned Poisson statistic ($C$), Pearson's ${\chi }^2$ and Neyman's ${\chi }_N^2$, the Kolmogorov- Smirnov, and Kuiper' statistics. It is found that the unbinned log likelihood $L$ performs best with regard to the expected chi-square distance between true and estimated spectra, the chance of a successful identification among discrete candidate models, the area under the receiver-operator curve of reduced (two-model) binary classification problems, and generally also with regard to the mean square errors of individual spectrum parameters. The ${\chi }^2$ (${\chi }_{\mathrm{N}}^2$) statistics should only be used if more than 10 (15) predicted counts per bin are available. From the practical point of view, the computational cost of evaluating $L$ is smaller than for any of the alternative methods if the forward model is specified in terms of a Poisson intensity and normalization is a free parameter. The maximum-$L$ method is applied to 14 observations from the Taurus Molecular Cloud, and the unbinned results are compared to binned XSPEC results, and found to generally agree, with exceptions explained by instability under re-binning and by background fine structures. The maximum-$L$ method has no lower limit on the available counts, and allows to treat weak sources which are beyond the means of binned methods.

accepted for publication in A&A

References in corpus (18)

• A Unified Approach to the Classical Statistical Analysis of Small Signals
• Statistics: Handle with Care, Detecting Multiple Model Components with the Likelihood Ratio Test
• Observational Constraints on the Formation and Evolution of Binary Stars
• X-ray astronomy of stellar coronae (Review)
• Bayesian Blocks, A New Method to Analyze Structure in Photon Counting Data
• Evidence for Accretion: High-Resolution X-ray Spectroscopy of the Classical T Tauri Star TW Hydrae
• The XMM-Newton Extended Survey of the Taurus Molecular Cloud (XEST)
• Including Systematic Uncertainties in Confidence Interval Construction for Poisson Statistics
• Determination of the gas-to-dust ratio in nearby dense clouds using X-ray absorption measurements
• New Brown Dwarfs and an Updated Initial Mass Function in Taurus
• Coronal Evolution of the Sun in Time: High-Resolution X-Ray Spectroscopy of Solar Analogs with Different Ages
• Parameter Estimation in Astronomy with Poisson-Distributed Data. I. The Chi-Square-Gamma Statistic
• Evidence for an X-Ray Jet in DG Tau A?
• X-rays from jet-driving protostars and T Tauri stars
• Coronal properties of G-type stars in different evolutionary phases
• The Coronae of AB Dor and V471 Tau: Primordial angular momentum vs tidal spin-up
• A new ordering principle for the classical statistical analysis of Poisson processes with background
• A two-dimensional Kolmogorov-Smirnov test for crowded field source detection: ROSAT sources in NGC 6397

Cited by in corpus (7)

• The XMM-Newton Extended Survey of the Taurus Molecular Cloud (XEST)
• Chi-square and Poissonian Data: Biases Even in the High-Count Regime and How to Avoid them
• Dark matter line emission constraints from NuSTAR observations of the Bullet Cluster
• Mass Outflow of the X-ray Emission Line Gas in NGC 4151
• The ALHAMBRA survey: an empirical estimation of the cosmic variance for merger fraction studies based on close pairs
• Stellar populations of galaxies in the ALHAMBRA survey up to $z\sim 1$. II. Stellar content of quiescent galaxies within the dust-corrected stellar mass$-$colour and the $UVJ$ colour$-$colour diagrams
• Stellar populations of galaxies in the ALHAMBRA survey up to $z\sim 1$. III. The stellar content of the quiescent galaxy population during the last $8$ Gyr