arXiv subjects

E. Cameron

Publications and source records attributed to E. Cameron.

At least 19 recordsLinked to original sources

arXiv2020

Women in academia: a warning on selection bias in gender studies from the astronomical perspective

The recent paper by AlShebli et al. (2020) investigates the impact of mentorship in young scientists. Among their conclusions, they state that female prot\'eg\'es benefit more from male than female mentorship. We herein expose a critical flaw in their methodological design that is a common issue in Astronomy, namely "selection biases". An effect that if not treated properly may lead to unwarranted causality claims. In their analysis, selection biases seem to be present in the response rate of their survey (8.35%), the choice of database, success criterion, and the overlook of the numerous drawbacks female researchers face in academia. We discuss these issues and their implications -- one of them being the potential increase in obstacles for women in academia. Finally, we reinforce the dangers of not considering selection bias effects in studies aimed at retrieving causal relations.

physics.soc-ph↗

arXiv2018

Spatial field reconstruction with INLA: Application to IFU galaxy data

Astronomical observations of extended sources, such as cubes of integral field spectroscopy (IFS), encode auto-correlated spatial structures that cannot be optimally exploited by standard methodologies. This work introduces a novel technique to model IFS datasets, which treats the observed galaxy properties as realizations of an unobserved Gaussian Markov random field. The method is computationally efficient, resilient to the presence of low-signal-to-noise regions, and uses an alternative to Markov Chain Monte Carlo for fast Bayesian inference, the Integrated Nested Laplace Approximation (INLA). As a case study, we analyse 721 IFS data cubes of nearby galaxies from the CALIFA and PISCO surveys, for which we retrieve the maps of the following physical properties: age, metallicity, mass and extinction. The proposed Bayesian approach, built on a generative representation of the galaxy properties, enables the creation of synthetic images, recovery of areas with bad pixels, and an increased power to detect structures in datasets subject to substantial noise and/or sparsity of sampling. A snippet code to reproduce the analysis of this paper is available in the COIN toolbox, together with the field reconstructions of the CALIFA and PISCO samples.

astro-ph.IM↗

arXiv2016

Is the cluster environment quenching the Seyfert activity in elliptical and spiral galaxies?

We developed a hierarchical Bayesian model (HBM) to investigate how the presence of Seyfert activity relates to their environment, herein represented by the galaxy cluster mass, $M_{200}$, and the normalized cluster-centric distance, $r/r_{200}$. We achieved this by constructing an unbiased sample of galaxies from the Sloan Digital Sky Survey, with morphological classifications provided by the Galaxy Zoo Project. A propensity score matching approach is introduced to control for the effects of confounding variables: stellar mass, galaxy colour, and star formation rate. The connection between Seyfert-activity and environmental properties in the de-biased sample is modelled within an HBM framework using the so-called logistic regression technique, suitable for the analysis of binary data (e.g., whether or not a galaxy hosts an AGN). Unlike standard ordinary least square fitting methods, our methodology naturally allows modelling the probability of Seyfert-AGN activity in galaxies on their natural scale, i.e. as a binary variable. Furthermore, we demonstrate how an HBM can incorporate information of each particular galaxy morphological type in a unified framework. In elliptical galaxies, our analysis indicates a strong correlation of Seyfert-AGN activity with $r/r_{200}$, and a weaker correlation with the mass of the host. In spiral galaxies these trends do not appear, suggesting that the link between Seyfert activity and the properties of spiral galaxies are independent of the environment.

astro-ph.GA↗

arXiv2015

Using gamma regression for photometric redshifts of survey galaxies

Machine learning techniques offer a plethora of opportunities in tackling big data within the astronomical community. We present the set of Generalized Linear Models as a fast alternative for determining photometric redshifts of galaxies, a set of tools not commonly applied within astronomy, despite being widely used in other professions. With this technique, we achieve catastrophic outlier rates of the order of ~1%, that can be achieved in a matter of seconds on large datasets of size ~1,000,000. To make these techniques easily accessible to the astronomical community, we developed a set of libraries and tools that are publicly available.

astro-ph.IM↗

arXiv2015

The Overlooked Potential of Generalized Linear Models in Astronomy-III: Bayesian Negative Binomial Regression and Globular Cluster Populations

In this paper, the third in a series illustrating the power of generalized linear models (GLMs) for the astronomical community, we elucidate the potential of the class of GLMs which handles count data. The size of a galaxy's globular cluster population $N_{\rm GC}$ is a prolonged puzzle in the astronomical literature. It falls in the category of count data analysis, yet it is usually modelled as if it were a continuous response variable. We have developed a Bayesian negative binomial regression model to study the connection between $N_{\rm GC}$ and the following galaxy properties: central black hole mass, dynamical bulge mass, bulge velocity dispersion, and absolute visual magnitude. The methodology introduced herein naturally accounts for heteroscedasticity, intrinsic scatter, errors in measurements in both axes (either discrete or continuous), and allows modelling the population of globular clusters on their natural scale as a non-negative integer variable. Prediction intervals of 99% around the trend for expected $N_{\rm GC}$comfortably envelope the data, notably including the Milky Way, which has hitherto been considered a problematic outlier. Finally, we demonstrate how random intercept models can incorporate information of each particular galaxy morphological type. Bayesian variable selection methodology allows for automatically identifying galaxy types with different productions of GCs, suggesting that on average S0 galaxies have a GC population 35% smaller than other types with similar brightness.

astro-ph.IM↗

arXiv2015

cosmoabc: Likelihood-free inference via Population Monte Carlo Approximate Bayesian Computation

Approximate Bayesian Computation (ABC) enables parameter inference for complex physical systems in cases where the true likelihood function is unknown, unavailable, or computationally too expensive. It relies on the forward simulation of mock data and comparison between observed and synthetic catalogues. Here we present cosmoabc, a Python ABC sampler featuring a Population Monte Carlo (PMC) variation of the original ABC algorithm, which uses an adaptive importance sampling scheme. The code is very flexible and can be easily coupled to an external simulator, while allowing to incorporate arbitrary distance and prior functions. As an example of practical application, we coupled cosmoabc with the numcosmo library and demonstrate how it can be used to estimate posterior probability distributions over cosmological parameters based on measurements of galaxy clusters number counts without computing the likelihood function. cosmoabc is published under the GPLv3 license on PyPI and GitHub and documentation is available at http://goo.gl/SmB8EX

astro-ph.CO↗

arXiv2014

The Overlooked Potential of Generalized Linear Models in Astronomy-II: Gamma regression and photometric redshifts

Machine learning techniques offer a precious tool box for use within astronomy to solve problems involving so-called big data. They provide a means to make accurate predictions about a particular system without prior knowledge of the underlying physical processes of the data. In this article, and the companion papers of this series, we present the set of Generalized Linear Models (GLMs) as a fast alternative method for tackling general astronomical problems, including the ones related to the machine learning paradigm. To demonstrate the applicability of GLMs to inherently positive and continuous physical observables, we explore their use in estimating the photometric redshifts of galaxies from their multi-wavelength photometry. Using the gamma family with a log link function we predict redshifts from the PHoto-z Accuracy Testing simulated catalogue and a subset of the Sloan Digital Sky Survey from Data Release 10. We obtain fits that result in catastrophic outlier rates as low as ~1% for simulated and ~2% for real data. Moreover, we can easily obtain such levels of precision within a matter of seconds on a normal desktop computer and with training sets that contain merely thousands of galaxies. Our software is made publicly available as an user-friendly package developed in Python, R and via an interactive web application (https://cosmostatisticsinitiative.shinyapps.io/CosmoPhotoz). This software allows users to apply a set of GLMs to their own photometric catalogues and generates publication quality plots with minimum effort from the user. By facilitating their ease of use to the astronomical community, this paper series aims to make GLMs widely known and to encourage their implementation in future large-scale projects, such as the Large Synoptic Survey Telescope.

astro-ph.IM↗

arXiv2014

The Overlooked Potential of Generalized Linear Models in Astronomy - I: Binomial Regression

Revealing hidden patterns in astronomical data is often the path to fundamental scientific breakthroughs; meanwhile the complexity of scientific inquiry increases as more subtle relationships are sought. Contemporary data analysis problems often elude the capabilities of classical statistical techniques, suggesting the use of cutting edge statistical methods. In this light, astronomers have overlooked a whole family of statistical techniques for exploratory data analysis and robust regression, the so-called Generalized Linear Models (GLMs). In this paper -- the first in a series aimed at illustrating the power of these methods in astronomical applications -- we elucidate the potential of a particular class of GLMs for handling binary/binomial data, the so-called logit and probit regression techniques, from both a maximum likelihood and a Bayesian perspective. As a case in point, we present the use of these GLMs to explore the conditions of star formation activity and metal enrichment in primordial minihaloes from cosmological hydro-simulations including detailed chemistry, gas physics, and stellar feedback. We predict that for a dark mini-halo with metallicity $\approx 1.3 \times 10^{-4} Z_{\bigodot}$, an increase of $1.2 \times 10^{-2}$ in the gas molecular fraction, increases the probability of star formation occurrence by a factor of 75%. Finally, we highlight the use of receiver operating characteristic curves as a diagnostic for binary classifiers, and ultimately we use these to demonstrate the competitive predictive performance of GLMs against the popular technique of artificial neural networks.

astro-ph.IM↗

arXiv2014

Galaxy And Mass Assembly (GAMA): Testing galaxy formation models through the most massive galaxies in the Universe

We have analysed the growth of Brightest Group Galaxies and Brightest Cluster Galaxies (BGGs/BCGs) over the last 3 billion years using a large sample of 883 galaxies from the Galaxy And Mass Assembly Survey. By comparing the stellar mass of BGGs and BCGs in groups and clusters of similar dynamical masses, we find no significant growth between redshift $z=0.27$ and $z=0.09$. We also examine the number of BGGs/BCGs that have line emission, finding that approximately 65 per cent of BGGs/BCGs show H$α$ in emission. From the galaxies where the necessary spectroscopic lines were accurately recovered (54 per cent of the sample), we find that half of this (i.e. 27 per cent of the sample) harbour on-going star formation with rates up to $10\,$M$_{\odot}$yr$^{-1}$, and the other half (i.e. 27 per cent of the sample) have an active nucleus (AGN) at the centre. BGGs are more likely to have ongoing star formation, while BCGs show a higher fraction of AGN activity. By examining the position of the BGGs/BCGs with respect to their host dark matter halo we find that around 13 per cent of them do not lie at the centre of the dark matter halo. This could be an indicator of recent cluster-cluster mergers. We conclude that BGGs and BCGs acquired their stellar mass rapidly at higher redshifts as predicted by semi-analytic models, mildly slowing down at low redshifts.

astro-ph.CO↗

arXiv2013

Importance Nested Sampling and the MultiNest Algorithm

Bayesian inference involves two main computational challenges. First, in estimating the parameters of some model for the data, the posterior distribution may well be highly multi-modal: a regime in which the convergence to stationarity of traditional Markov Chain Monte Carlo (MCMC) techniques becomes incredibly slow. Second, in selecting between a set of competing models the necessary estimation of the Bayesian evidence for each is, by definition, a (possibly high-dimensional) integration over the entire parameter space; again this can be a daunting computational task, although new Monte Carlo (MC) integration algorithms offer solutions of ever increasing efficiency. Nested sampling (NS) is one such contemporary MC strategy targeted at calculation of the Bayesian evidence, but which also enables posterior inference as a by-product, thereby allowing simultaneous parameter estimation and model selection. The widely-used MultiNest algorithm presents a particularly efficient implementation of the NS technique for multi-modal posteriors. In this paper we discuss importance nested sampling (INS), an alternative summation of the MultiNest draws, which can calculate the Bayesian evidence at up to an order of magnitude higher accuracy than `vanilla' NS with no change in the way MultiNest explores the parameter space. This is accomplished by treating as a (pseudo-)importance sample the totality of points collected by MultiNest, including those previously discarded under the constrained likelihood sampling of the NS algorithm. We apply this technique to several challenging test problems and compare the accuracy of Bayesian evidences obtained with INS against those from vanilla NS.

astro-ph.IM↗

arXiv2013

Newly-quenched galaxies as the cause for the apparent evolution in average size of the population

Abridged. We use COSMOS to study in a self-consistent way the change in the number densities of quenched early-type galaxies (Q-ETGs) of a given size over the interval 0.2 < z < 1.0 to study the claimed size evolution of these galaxies. At 10^10.5 10^11 Msun we find a decrease by 30%. In both mass bins, the increase of the median sizes of Q-ETGs with time is primarily caused by the addition to the size function of larger and more diffuse Q-ETGs. At all masses, compact Q-ETGs become systematically redder towards later epochs, with a (U-V) difference consistent with passive evolution of their stellar populations, indicating that they are a population that does not appreciably evolve in size. At all epochs, the larger Q-ETGs (at least in the lower mass bin) have average rest-frame colors systematically bluer than those of the more compact Q-ETGs, suggesting that the former are younger than the latter. The idea that new, large, Q-ETGs are responsible for the observed growth in the median size of the population at a given mass is supported by the sizes and number of the star-forming galaxies that are expected to be progenitors of the new Q-ETGs over the same period. In the low mass bin, the new Q-ETG have 30% smaller sizes than their star-forming progenitors. This is likely due to the fading of their disks after they cease star-formation. Comparison with higher z shows that the median size of newly-quenched galaxies roughly scales, at constant mass, as (1+z)^-1. The dominant cause of the size evolution seen in the Q-ETG population is thus that the average sizes of individual Q-ETGs scale with the average density of the Universe at the time when they were quenched, with subsequent size changes in individual objects through eg merging of secondary importance, especially at masses <10^11 Msun.

astro-ph.CO↗

arXiv2013

Galaxy And Mass Assembly (GAMA): Spectroscopic analysis

The Galaxy And Mass Assembly (GAMA) survey is a multiwavelength photometric and spectroscopic survey, using the AAOmega spectrograph on the Anglo-Australian Telescope to obtain spectra for up to ~300000 galaxies over 280 square degrees, to a limiting magnitude of r_pet < 19.8 mag. The target galaxies are distributed over 0<z<0.5 with a median redshift of z~0.2, although the redshift distribution includes a small number of systems, primarily quasars, at higher redshifts, up to and beyond z=1. The redshift accuracy ranges from sigma_v~50km/s to sigma_v~100km/s depending on the signal-to-noise of the spectrum. Here we describe the GAMA spectroscopic reduction and analysis pipeline. We present the steps involved in taking the raw two-dimensional spectroscopic images through to flux-calibrated one-dimensional spectra. The resulting GAMA spectra cover an observed wavelength range of 3750<lambda<8850 A at a resolution of R~1300. The final flux calibration is typically accurate to 10-20%, although the reliability is worse at the extreme wavelength ends, and poorer in the blue than the red. We present details of the measurement of emission and absorption features in the GAMA spectra. These measurements are characterised through a variety of quality control analyses detailing the robustness and reliability of the measurements. We illustrate the quality of the measurements with a brief exploration of elementary emission line properties of the galaxies in the GAMA sample. We demonstrate the luminosity dependence of the Balmer decrement, consistent with previously published results, and explore further how Balmer decrement varies with galaxy mass and redshift. We also investigate the mass and redshift dependencies of the [NII]/Halpha vs [OIII]/Hbeta spectral diagnostic diagram, commonly used to discriminate between star forming and nuclear activity in galaxies.

astro-ph.CO↗

arXiv2012

Herschel-ATLAS/GAMA: spatial clustering of low-redshift sub-mm galaxies

We have measured the clustering properties of low-redshift (z < 0.3) sub-mm galaxies detected at 250 micron in the Herschel-ATLAS Science Demonstration Phase (SDP) field. We selected a sample for which we have high-quality spectroscopic redshifts, obtained from reliably matching the 250-micron sources to a complete (for r < 19.4) sample of galaxies from the GAMA database. Both the angular and spatial clustering strength are measured for all z < 0.3 sources as well as for five redshift slices with thickness delta z=0.05 in the range 0.05 < z < 0.3. Our measured spatial clustering length r_0 is comparable to that of optically-selected, moderately star-forming (blue) galaxies: we find values around 5 Mpc. One of the redshift bins contains an interesting structure, at z = 0.164.

astro-ph.CO↗

arXiv2012

Galaxy And Mass Assembly (GAMA): The mass-metallicity relationship

Context: The mass-metallicity relationship (MMR) of star-forming galaxies is well-established, however there is still some disagreement with respect to its exact shape and its possible dependence on other observables. Aims: We measure the MMR in the Galaxy And Mass Assembly (GAMA) survey. We compare our measured MMR to that measured in the Sloan Digital Sky Survey (SDSS) and study the dependence of the MMR on various selection criteria to identify potential causes for disparities seen in the literature. Methods: We use strong emission line ratio diagnostics to derive oxygen abundances. We then apply a range of selection criteria for the minimum signal-to-noise in various emission lines, as well as the apparent and absolute magnitude to study variations in the inferred MMR. Results: The shape and position of the MMR can differ significantly depending on the metallicity calibration and selection used. After selecting a robust metallicity calibration amongst those tested, we find that the mass-metallicity relation for redshifts 0.061< z<0.35 in GAMA is in reasonable agreement with that found in the SDSS despite the difference in the luminosity range probed. Conclusions: In view of the significant variations of the MMR brought about by reasonable changes in the sample selection criteria and method, we recommend that care be taken when comparing the MMR from different surveys and studies directly. We also conclude that there could be a modest level of evolution over 0.06<z<0.35 within the GAMA sample.

astro-ph.CO↗

arXiv2012

Galaxy And Mass Assembly (GAMA): The 0.013 < z < 0.1 cosmic spectral energy distribution from 0.1 micron to 1mm

We use the GAMA I dataset combined with GALEX, SDSS and UKIDSS imaging to construct the low-redshift (z<0.1) galaxy luminosity functions in FUV, NUV, ugriz, and YJHK bands from within a single well constrained volume of 3.4 x 10^5 (Mpc/h)^{3}. The derived luminosity distributions are normalised to the SDSS DR7 main survey to reduce the estimated cosmic variance to the 5 per cent level. The data are used to construct the cosmic spectral energy distribution (CSED) from 0.1 to 2.1 \mum free from any wavelength dependent cosmic variance for both the elliptical and non-elliptical populations. The two populations exhibit dramatically different CSEDs as expected for a predominantly old and young population respectively. Using the Driver et al. (2008) prescription for the azimuthally averaged photon escape fraction, the non-ellipticals are corrected for the impact of dust attenuation and the combined CSED constructed. The final results show that the Universe is currently generating (1.8 +/- 0.3) x 10^{35} h W Mpc^{-3} of which (1.2 +/- 0.1) x 10^{35} h W Mpc^{-3} is directly released into the inter-galactic medium and (0.6 +/- 0.1) x 10^{35} h W Mpc^{-3} is reprocessed and reradiated by dust in the far-IR. Using the GAMA data and our dust model we predict the mid and far-IR emission which agrees remarkably well with available data. We therefore provide a robust description of the pre- and post dust attenuated energy output of the nearby Universe from 0.1micron to 0.6mm. The largest uncertainty in this measurement lies in the mid and far-IR bands stemming from the dust attenuation correction and its currently poorly constrained dependence on environment, stellar mass, and morphology.

astro-ph.CO↗

arXiv2012

The Zurich Environmental Study of Galaxies in Groups along the Cosmic Web. III. Galaxy Photometric Measurements and the Spatially-Resolved Color Properties of Early- and Late-Type Satellites in Diverse Environments

We present photometric measurements for the galaxies - and when possible their bulges and disks - in the 0.05 10^10 Msun, these satellites have shallower color gradients in higher mass groups and in the cores of groups compared with lower mass groups and the outskirts of groups. Stellar population analyses and semi-analytic models suggest that disk-dominated satellites undergo quenching of star formation in their outer disks, on timescales ~2 Gyr, as they progressively move inside the group potential.

astro-ph.CO↗

arXiv2012

The Zurich Environmental Study (ZENS) of Galaxies in Groups along the Cosmic Web. II. Galaxy Structural Measurements and the Concentration of Morphologically Classified Satellites in Diverse Environments

We present structural measurements for the galaxies in the 0.05 10^10 M_sun, the concentration of disk satellites increases with increasing stellar mass, separately within each morphological bin of B/T. The known increase in concentration with stellar mass for disk satellites is thus due, at least in part, to an increase in galaxy central stellar density at constant B/T. The correlation between concentration and galaxy stellar mass becomes progressively steeper for later morphological types. The concentration of disk satellites shows a barely significant dependence on delta_LSS or R/R_200. The strongest environmental effect is found with group mass for M>10^10 M_sun disk-dominated satellites, which are ~10% more concentrated in high mass groups than in lower mass groups.

astro-ph.CO↗

arXiv2012

The Zurich Environmental Study (ZENS) of Galaxies in Groups along the Cosmic Web. I. Which Environment Affects Galaxy Evolution?

The Zurich Environmental Study (ZENS) is based on a sample of ~1500 galaxy members of 141 groups in the mass range ~10^12.5-14.5 M_sun within the narrow redshift range 0.05 10^10 M_sun, satellite galaxies in relaxed and unrelaxed groups have similar size, color and (specific) star formation rate distributions; at lower galaxy masses, satellites are marginally redder in relaxed relative to unrelaxed groups, suggesting quenching of star formation in low-mass satellites by physical processes active in relaxed halos. Finally, relaxed and unrelated groups show similar stellar mass conversion efficiencies, peaking at halo masses around 10^12.5 M_sun. In the enclosed ZENS catalogue we publish all environmental diagnostics as well as the galaxy structural and photometric measurements described in companion ZENS papers II and III.

astro-ph.CO↗