Search

The homologous collapse from rest of a uniform density sphere under its self gravity is a well-known toy model for the formation dynamics of astronomical objects ranging from stars to galaxies. Equally well-known is that the evolution of the radius with time cannot be explicitly obtained because of the transcendental nature of the differential equation solution. Rather, both radius and time are written parametrically in terms of the development angle $\theta$. We here present an explicit integral solution for radius as a function of time, exploiting methods from complex analysis recently applied to the mathematically-similar 'geometric goat problem'. Our solution can be efficiently evaluated using a Fast Fourier Transform and allows for arbitrary sampling in time, with a simple Python implementation that is $\sim$$100\times$ faster than using numerical root-finding to achieve arbitrary sampling. Our explicit solution is advantageous relative to the usual approach of first generating a uniform grid in $\theta$, since this latter results in a non-uniform radial or time sampling, less useful for applications such as generation of sub-grid physics models., Comment: 4 pages, 3 figures, submitted to MNRAS. Python implementation available at https://gist.github.com/oliverphilcox/559f086f1bf63b23d55c508b2f47bad3

The widely used MASTER approach for angular power spectrum estimation was developed as a fast $C_{\ell}$ estimator on limited regions of the sky. This method expresses the power spectrum of a masked map ("pseudo-$C_\ell$") in terms of the power spectrum of the unmasked map (the true $C_\ell$) and that of the mask or weight map. However, it is often the case that the map and mask are correlated in some way, such as point source masks used in cosmic microwave background (CMB) analyses, which have nonzero correlation with CMB secondary anisotropy fields and other mm-wave sky signals. In such situations, the MASTER approach gives biased results, as it assumes that the unmasked map and mask have zero correlation. While such effects have been discussed before with regard to specific physical models, here we derive a completely general formalism for any case where the map and mask are correlated. We show that our result ("reMASTERed") reconstructs ensemble-averaged pseudo-$C_\ell$ to effectively exact precision, with significant improvements over traditional estimators for cases where the map and mask are correlated. In particular, we obtain an improvement in the mean absolute percent error from 30% with the MASTER result to essentially no error with the reMASTERed result for an integrated Sachs-Wolfe (ISW) field map with a mask built from the thresholded ISW field, and 10% to effectively zero for a Compton-$y$ map combined with an infrared source mask (the latter being directly relevant to actual data analysis). An important consequence of our result is that for maps with correlated masks it is no longer possible to invert a simple equation to obtain the true $C_\ell$ from the pseudo-$C_\ell$. Instead, our result necessitates the use of forward modeling from theory space into the observable domain of the pseudo-$C_\ell$. Our code is publicly available at https://github.com/kmsurrao/reMASTERed., 14+9 pages, 6+6 figures; matches the version accepted for publication in PRD (https://journals.aps.org/prd/abstract/10.1103/PhysRevD.107.083521); code available at at https://github.com/kmsurrao/reMASTERed

We present efficient algorithms for computing the

We present a framework for self-consistent cosmological analyses of the full-shape anisotropic bispectrum, including the quadrupole $(\ell=2)$ and hexadecapole $(\ell=4)$ moments. This features a novel window-free algorithm for extracting the latter quantities from data, derived using a maximum-likelihood prescription. Furthermore, we introduce a theoretical model for the bispectrum multipoles (which does not introduce new free parameters), and test both aspects of the pipeline on several high-fidelity mocks, including the PT Challenge suite of gigantic cumulative volume. This establishes that the systematic error is significantly below the statistical threshold, both for the measurement and modeling. As a realistic example, we extract the large-scale bispectrum multipoles from BOSS DR12 and analyze them in combination with the power spectrum data. Assuming a minimal $\Lambda$CDM model, with a BBN prior on the baryon density and a \textit{Planck} prior on $n_s$, we can extract the remaining cosmological parameters directly from the clustering data. The inclusion of the unwindowed higher-order $(\ell>0)$ large-scale bispectrum multipoles is found to moderately improve one-dimensional cosmological parameter posteriors (at the $5\%-10\%$ level), though these multipoles are detected only in three out of four BOSS data segments at $\approx 5\sigma$. Combining information from the power spectrum and bispectrum multipoles, the real space power spectrum, and the post-reconstructed BAO data, we find $H_0 = 68.2\pm 0.8~\mathrm{km}\,\mathrm{s}^{-1}\mathrm{Mpc}^{-1}$, $\Omega_m =0.33\pm 0.01$ and $\sigma_8 = 0.736\pm 0.033$ (the tightest yet found in perturbative full-shape analyses). Our estimate of the growth parameter $S_8=0.77\pm 0.04$ agrees with both weak lensing and CMB results., Comment: 40 pages, 8 figures, 6 tables; estimators and data are publicly available at https://github.com/oliverphilcox/Spectra-Without-Windows

Could new physics break the mirror symmetry of the Universe? Utilizing recent measurements of the parity-odd four-point correlation function of BOSS galaxies, we probe the physics of inflation by placing constraints on the amplitude of a number of parity-violating models. Within canonical models of (single-field, slow-roll) inflation, no parity-asymmetry can occur; however, it has recently been shown that breaking of the standard assumptions can lead to parity violation within the Effective Field Theory of Inflation (EFTI). In particular, we consider the Ghost Condensate and Cosmological Collider scenarios -- the former for the leading and subleading operators in the EFTI and the latter for different values of mass and speed of an exchanged spin-$1$ particle -- for a total of $18$ models. Each instance yields a definite prediction for the inflationary trispectrum, which we convert to a late-time galaxy correlator prediction (through a highly non-trivial calculation) and constrain using the observed data. We find no evidence for inflationary parity-violation (with each of the $18$ models having significances below $2\sigma$), and place the first constraints on the relevant coupling strengths, at a level comparable with the theoretical perturbativity bounds. This is also the first time Cosmological Collider signatures have directly been searched for in observational data. We further show that possible secondary parity-violating signatures in galaxy clustering can be systematically described within the Effective Field Theory of Large-Scale Structure. We argue that these late-time contributions are subdominant compared to the primordial parity-odd signal for a vast region of parameter space. In summary, the results of this paper disfavor the notion that the recent hints of parity-violation observed in the distribution of galaxies are due to new physics., Comment: 20+9 spooky pages, 4 figures, accepted by Phys. Rev. D. Code available at https://github.com/oliverphilcox/Parity-Odd-4PCF

We present a new algorithm for efficiently computing the N-point correlation functions (NPCFs) of a 3D density field for arbitrary N. This can be applied both to a discrete spectroscopic galaxy survey and a continuous field. By expanding the statistics in a separable basis of isotropic functions built from spherical harmonics, the NPCFs can be estimated by counting pairs of particles in space, leading to an algorithm with complexity $\mathcal {O}(N_\mathrm{g}^2)$ for Ng particles, or $\mathcal {O}(N_\mathrm{FFT}\log N_\mathrm{FFT})$ when using a Fast Fourier Transform with NFFT grid-points. In practice, the rate-limiting step for N > 3 will often be the summation of the histogrammed spherical harmonic coefficients, particularly if the number of radial and angular bins is large. In this case, the algorithm scales linearly with Ng. The approach is implemented in the encore code, which can compute the 3PCF, 4PCF, 5PCF, and 6PCF of a BOSS-like galaxy survey in ${\sim}100$ CPU-hours, including the corrections necessary for non-uniform survey geometries. We discuss the implementation in depth, along with its GPU acceleration, and provide practical demonstration on realistic galaxy catalogues. Our approach can be straightforwardly applied to current and future data sets to unlock the potential of constraining cosmology from the higher point functions.

The polarized Sunyaev-Zel'dovich (pSZ) effect is sourced by the Thomson scattering of CMB photons from distant free electrons and yields a novel view of the CMB quadrupole throughout the observable Universe. Galaxy shear measures the shape distortions of galaxies, probing both their local environment and the intervening matter distribution. Both observables give interesting constraints on the cosmological model; here we ask: what can be learnt from their combination? The pSZ-shear cross-spectrum measures the shear-galaxy-polarization bispectrum and contains contributions from (1) the Sachs-Wolfe (SW) effect, (2) the integrated Sachs-Wolfe (ISW) effect, and (3) inflationary gravitational waves. Since the modes contributing to the pSZ signal are not restricted to the Earth's past lightcone, the low-redshift cross-spectra could provide a novel constraint on dark energy via the ISW effect, whilst the SW signal couples scalar modes at very different times but at similar positions; this provides a unique probe of the Universe's homogeneous time evolution. We give expressions for all major contributions to the shear, galaxy, and pSZ auto- and cross-spectra, and evaluate their detectability via Fisher forecasts. Despite significant theoretical utility, the cross-spectra will be challenging to detect: combining CMB-S4 with Rubin yields a $1.6\sigma$ detection of the ISW contribution, which increases to $5.2\sigma$ for a futuristic experiment involving CMB-HD and a higher galaxy sample density. For parity-even (parity-odd) tensors, we predict a $1\sigma$ limit of $\sigma(r) = 0.9$ ($0.2$) for CMB-S4 and Rubin, or $0.3$ ($0.06$) for the more futuristic setup. Whilst this is significantly better than the constraints from galaxy shear alone (and less sensitive to systematics), it is unlikely to be competitive, but may serve as a useful cross-check., Comment: 15+5 pages, accepted by Phys. Rev. D. Code available at https://github.com/oliverphilcox/pSZ-cross-Shear

We derive analytic covariance matrices for the $N$-Point Correlation Functions (NPCFs) of galaxies in the Gaussian limit. Our results are given for arbitrary $N$ and projected onto the isotropic basis functions of Cahn & Slepian (2020), recently shown to facilitate efficient NPCF estimation. A numerical implementation of the 4PCF covariance is compared to the sample covariance obtained from a set of lognormal simulations, Quijote dark matter halo catalogues, and MultiDark-Patchy galaxy mocks, with the latter including realistic survey geometry. The analytic formalism gives reasonable predictions for the covariances estimated from mock simulations with a periodic-box geometry. Furthermore, fitting for an effective volume and number density by maximizing a likelihood based on Kullback-Leibler divergence is shown to partially compensate for the effects of a non-uniform window function., 31 pages, 17 figures

We use redshift-space galaxy clustering data from the BOSS survey to constrain local primordial non-Gaussianity (LPNG). This is of particular importance due to the consistency relations, which imply that a detection of LPNG would rule out all single-field inflationary models. Our constraints are based on the consistently analyzed redshift-space galaxy power spectra and bispectra, extracted from the public BOSS data with optimal window-free estimators. We use a complete perturbation theory model including all one-loop power spectrum corrections generated by LPNG. Our constraint on the amplitude of the local non-Gaussian shape is $f_{\rm NL}^{\rm local}=-33\pm 28$ at 68\%\,CL, yielding no evidence for primordial non-Gaussianity. The addition of the bispectrum tightens the $f_{\rm NL}^{\rm local}$ constraints from BOSS by $20\%$, and allows breaking of degeneracies with non-Gaussian galaxy bias. These results set the stage for the analysis of future surveys, whose larger volumes will yield significantly tighter constraints on LPNG., 20 pages, 5 figures

We show how consistency relations can be used to robustly extract the amplitude of local primordial non-Gaussianity ($f_{\rm NL}$) from the squeezed limit of the matter bispectrum, well into the non-linear regime. First, we derive a non-perturbative relation between primordial non-Gaussianity and the leading term in the squeezed bispectrum, revising some results present in the literature. This relation is then used to successfully measure $f_{\rm NL}$ from $N$-body simulations. We discuss the dependence of our results on different scale cuts and redshifts. Specifically, the analysis is strongly dependent on the choice of the smallest soft momentum, $q_{\rm min}$, which is the most sensitive to primordial bispectrum contributions, but is largely independent of the choice of the largest hard momentum, $k_{\rm max}$, due to the non-Gaussian nature of the covariance. We also show how the constraints on $f_{\rm NL}$ improve at higher redshift, due to a reduced off-diagonal covariance. In particular, for a simulation with $f_{\rm NL} = 100$ and a volume of $(2.4 \text{ Gpc}/h)^3$, we measure $f_{\rm NL} = 98 \pm 12$ at redshift $z=0$ and $f_{\rm NL} = 97 \pm 8$ at $z=0.97$. Finally, we compare our results with a Fisher forecast, showing that the current version of the analysis is satisfactorily close to the Fisher error. We regard this as a first step towards the realistic application of consistency relations to constrain primordial non-Gaussianity using observations., Comment: 17 pages, 8 figures. Minor changes. Matches version published in PRD

We study the information content of the angle-averaged (monopole) redshift space galaxy bispectrum. The main novelty of our approach is the use of a systematic tree-level perturbation theory model that includes galaxy bias, IR resummation, and also accounts for nonlinear redshift space distortions, binning, and projection effects. We analyze data from the PT challenge simulations, whose cumulative volume of 566 $h^{-3}$Gpc$^3$ allows for a precise comparison to theoretical predictions. Fitting the power spectrum and bispectrum of our simulated data, and varying all necessary cosmological and nuisance parameters in a consistent Markov chain Monte Carlo analysis, we find that our tree-level bispectrum model is valid up to $k_{\max}=0.08~h{\rm Mpc}^{-1}$ (at $z=0.61$). We also find that inclusion of the bispectrum monopole improves constraints on cosmological parameters by $(5-15)\%$ relative to the power spectrum. The improvement is more significant for the quadratic bias parameters of our simulated galaxies, which we also show to deviate from biases of the host dark matter halos at the $\sim 3\sigma$ level. Finally, we adjust the covariance and scale cuts to match the volume of the BOSS survey, and estimate that within the minimal $\Lambda$CDM model the bispectrum data can tighten the constraint on the mass fluctuation amplitude $\sigma_8$ by roughly $10\%$., Comment: 52 pages, 11 figures

Nonlinear redshift-space distortions (“fingers of God”) are challenging to model analytically, a fact that limits the applicability of perturbation theory (PT) in redshift space as compared to real space. We show how this problem can be mitigated using a new observable, Q0, which can be easily estimated from the redshift-space clustering data and is approximately equal to the real-space power spectrum. The new statistic does not suffer from fingers of God and can be accurately described with PT down to kmax≃0.4 h Mpc-1. It can be straightforwardly included in the likelihood at negligible additional computational cost and yields noticeable improvements on cosmological parameters compared to standard power spectrum multipole analyses. Using both simulations and observational data from the Baryon Oscillation Spectroscopic Survey, we show that improvements vary from 10% to 100% depending on the cosmological parameter considered, the galaxy sample, and the survey volume. Non-linear redshift-space distortions ("fingers of God") are challenging to model analytically, a fact that limits the applicability of perturbation theory in redshift space as compared to real space. We show how this problem can be mitigated using a new observable, $Q_0$, which can be easily estimated from the redshift space clustering data and is approximately equal to the real space power spectrum. The new statistic does not suffer from fingers of God and can be accurately described with perturbation theory down to $k_{\rm max}\simeq 0.4~h~\text{Mpc}^{-1}$. It can be straightforwardly included in the likelihood at negligible additional computational cost, and yields noticeable improvements on cosmological parameters compared to standard power spectrum multipole analyses. Using both simulations and observational data from the Baryon Oscillation Spectroscopic Survey, we show that improvements vary from $10\%$ to $100\%$ depending on the cosmological parameter considered, the galaxy sample and the survey volume.

To make use of clustering statistics from large cosmological surveys, accurate and precise covariance matrices are needed. We present a new code to estimate large scale galaxy two-point correlation function (2PCF) covariances in arbitrary survey geometries that, due to new sampling techniques, runs $\sim 10^4$ times faster than previous codes, computing finely-binned covariance matrices with negligible noise in less than 100 CPU-hours. As in previous works, non-Gaussianity is approximated via a small rescaling of shot-noise in the theoretical model, calibrated by comparing jackknife survey covariances to an associated jackknife model. The flexible code, RascalC, has been publicly released, and automatically takes care of all necessary pre- and post-processing, requiring only a single input dataset (without a prior 2PCF model). Deviations between large scale model covariances from a mock survey and those from a large suite of mocks are found to be be indistinguishable from noise. In addition, the choice of input mock are shown to be irrelevant for desired noise levels below $\sim 10^5$ mocks. Coupled with its generalization to multi-tracer data-sets, this shows the algorithm to be an excellent tool for analysis, reducing the need for large numbers of mock simulations to be computed., Comment: 29 pages, 8 figures. Accepted by MNRAS. Code is available at http://github.com/oliverphilcox/RascalC with documentation at http://rascalc.readthedocs.io/

We present configuration-space estimators for the auto- and cross-covariance of two- and three-point correlation functions (2PCF and 3PCF) in general survey geometries. These are derived in the Gaussian limit (setting higher-order correlation functions to zero), but for arbitrary non-linear 2PCFs (which may be estimated from the survey itself), with a shot-noise rescaling parameter included to capture non-Gaussianity. We generalize previous approaches to include Legendre moments via a geometry-correction function calibrated from measured pair and triple counts. Making use of importance sampling and random particle catalogs, we can estimate model covariances in fractions of the time required to do so with mocks, obtaining estimates with negligible sampling noise in $\sim 10$ ($\sim 100$) CPU-hours for the 2PCF (3PCF) auto-covariance. We compare results to sample covariances from a suite of BOSS DR12 mocks and find the matrices to be in good agreement, assuming a shot-noise rescaling parameter of $1.03$ ($1.20$) for the 2PCF (3PCF). To obtain strongest constraints on cosmological parameters we must use multiple statistics in concert; having robust methods to measure their covariances at low computational cost is thus of great relevance to upcoming surveys., Comment: 21 pages, 8 figures, accepted by MNRAS. Code is available at http://github.com/oliverphilcox/RascalC with documentation at http://rascalc.readthedocs.io/

Author(s): Philcox, OHE; Slepian, Z | Abstract: Conventional estimators of the anisotropic power spectrum and two-point correlation function (2PCF) adopt the "Yamamoto approximation,"fixing the line of sight of a pair of galaxies to that of just one of its members. While this is accurate only to first-order in the characteristic opening angle θmax, it allows for efficient implementation via fast Fourier transforms (FFTs). This work presents practical algorithms for computing the power spectrum and 2PCF multipoles using pairwise lines of sight, adopting either the galaxy midpoint or angle bisector definitions. Using newly derived infinite series expansions for spherical harmonics and Legendre polynomials, we construct estimators accurate to arbitrary order in θmax, though note that the midpoint and bisector formalisms themselves differ at fourth-order. Each estimator can be straightforwardly implemented using FFTs, requiring only modest additional computational cost relative to the Yamamoto approximation. We demonstrate the algorithms by applying them to a set of realistic mock galaxy catalogs, and find that both procedures produce comparable results for the 2PCF, with a slight preference for the bisector power spectrum algorithm, albeit at the cost of greater memory usage. Such estimators provide a useful method to reduce wide-angle systematics for future surveys.

Conventional algorithms for galaxy power-spectrum estimation measure the true spectrum convolved with a survey window function, which, for parameter inference, must be compared with a similarly convolved theory model. In this work, we directly estimate the unwindowed power spectrum multipoles using quadratic estimators akin to those introduced in the late 1990s. Under Gaussian assumptions, these are optimal and free from the leading-order effects of pixelization and non-Poissonian shot noise. They may be straightforwardly computed given the survey dataset and a suite of simulations of known cosmology. We implement the pixel-based maximum-likelihood estimator and a simplification based on the Feldman-Kaiser-Peacock weighting scheme, both of which can be computed via FFTs and conjugate-gradient descent methods. Furthermore, the estimators allow direct computation of spectrum coefficients in an arbitrary linear compression scheme, without needing to first bin the statistic. Applying the technique to a subset of the BOSS DR12 galaxies, we find that the pixel-based quadratic estimators give statistically consistent power spectra, compressed coefficients, and cosmological parameters to those obtained with the usual windowed approaches. Because of the sample's low number density and compact window function, the optimal weighting scheme gives little improvement over the simplified form; this may change for dense surveys or those focusing on primordial non-Gaussianity. The technique is shown to be efficient and robust, and shows significant potential for measuring the windowless power spectrum and bispectrum in the presence of weak non-Gaussianity.

A fundamental relation in celestial mechanics is Kepler's equation, linking an orbit's mean anomaly to its eccentric anomaly and eccentricity. Being transcendental, the equation cannot be directly solved for eccentric anomaly by conventional treatments; much work has been devoted to approximate methods. Here, we give an explicit integral solution, utilizing methods recently applied to the "geometric goat problem" and to the dynamics of spherical collapse. The solution is given as a ratio of contour integrals; these can be efficiently computed via numerical integration for arbitrary eccentricities. The method is found to be highly accurate in practice, with our C++ implementation outperforming conventional root-finding and series approaches by a factor greater than two., 7 pages, 4 figures, accepted by MNRAS. Code available at https://github.com/oliverphilcox/Keplers-Goat-Herd

A useful identity relating the infinite sum of two Bessel functions to their infinite integral was discovered in Dominici et al. (2012). Here, we extend this result to products of $N$ Bessel functions, and show it can be straightforwardly proven using the Abel-Plana theorem, or the Poisson summation formula. For $N=2$, the proof is much simpler than that of Dominici et al., and significantly enlarges the range of validity., Comment: 15 pages, 1 figure. Submitted to Proc. Roy. Soc. A

Two sources of geometric information are encoded in the galaxy power spectrum: the sound horizon at recombination and the horizon at matter-radiation equality. Analyzing the BOSS DR12 galaxy power spectra using perturbation theory with $\Omega_m$ priors from Pantheon supernovae but no priors on $\Omega_b$, we obtain constraints on $H_0$ from the second scale, finding $H_0 = 65.1^{+3.0}_{-5.4}\,\mathrm{km}\,\mathrm{s}^{-1}\mathrm{Mpc}^{-1}$; this differs from the best-fit of SH0ES at 95\% confidence. Similar results are obtained if $\Omega_m$ is constrained from uncalibrated BAO: $H_0 = 65.6^{+3.4}_{-5.5}\,\mathrm{km}\,\mathrm{s}^{-1}\mathrm{Mpc}^{-1}$. Adding the analogous lensing results from Baxter \& Sherwin 2020, the posterior shifts to $70.6^{+3.7}_{-5.0}\,\mathrm{km}\,\mathrm{s}^{-1}\mathrm{Mpc}^{-1}$. Using mock data, Fisher analyses, and scale-cuts, we demonstrate that our constraints do not receive significant information from the sound horizon scale. Since many models resolve the $H_0$ controversy by adding new physics to alter the sound horizon, our measurements are a consistency test for standard cosmology before recombination. A simple forecast indicates that such constraints could reach $\sigma_{H_0} \simeq 1.6\,\mathrm{km}\,\mathrm{s}^{-1}\mathrm{Mpc}^{-1}$ in the era of Euclid., Comment: 8 pages, 4 figures, accepted by Phys. Rev. D

We develop an analytical forward model based on perturbation theory to predict the redshift-space galaxy overdensity at the field level given a realization of the initial conditions. We find that the residual noise between the model and simulated galaxy density has a power spectrum that is white on large scales, with size comparable to the shot noise. In the mildly nonlinear regime, we see a $k^2\mu^2$ correction to the noise power spectrum, corresponding to larger noise along the line of sight and on smaller scales. The parametric form of this correction has been predicted on theoretical grounds before, and our simulations provide important confirmation of its presence. We have also modeled the galaxy velocity at the field-level and compared it against simulated galaxy velocities, finding that about $10\%$ of the galaxies are responsible for half of the rms velocity residual for our simulated galaxy sample., Comment: 21 pages, 12 figures. Code available at https://github.com/mschmittfull/perr

We perform a multi-tracer analysis using the complete Sloan Digital Sky Survey IV (SDSS-IV) extended Baryon Oscillation Spectroscopic Survey (eBOSS) DR16 luminous red galaxy (LRG) and the DR16 emission line galaxy (ELG) samples in the configuration space, and successfully detect a cross correlation between the two samples, and find the growth rate to be $f\sigma_8=0.342 \pm 0.085$ ($\sim25$ per cent accuracy) from the cross sample alone. We perform a joint measurement of the baryonic acoustic oscillation (BAO) and redshift space distortion (RSD) parameters at a single effective redshift of $z_{\rm eff}= 0.77$, using the auto- and cross-correlation functions of the LRG and ELG samples, and find that the comoving angular diameter distance $D_M(z_{\rm eff})/r_d = 18.85\pm 0.38$, the Hubble distance $D_H(z_{\rm eff})/r_d = 19.64 \pm 0.57$, and $f\sigma_8(z_{\rm eff}) = 0.432 \pm 0.038$, which is consistent with a $\Lambda$CDM model at $68\%$ CL. Compared to the single-tracer analysis on the LRG sample, the Figure of Merit (FoM) of $\alpha_{\perp}, \alpha_{||}$ and $f\sigma_8$ is improved by a factor of $1.11$ in our multi-tracer analysis, and in particular, the statistical uncertainty of $f\sigma_8$ is reduced by $11.6 \%$., Comment: 15 pages, 13 figures, 2 tables; MNRAS accepted; The BAO and RSD measurements and the covariance matrix are made available at https://github.com/ytcosmo/MultiTracerBAORSD/

We present the Cosmology and Astrophysics with MachinE Learning Simulations --CAMELS-- project. CAMELS is a suite of 4,233 cosmological simulations of $(25~h^{-1}{\rm Mpc})^3$ volume each: 2,184 state-of-the-art (magneto-)hydrodynamic simulations run with the AREPO and GIZMO codes, employing the same baryonic subgrid physics as the IllustrisTNG and SIMBA simulations, and 2,049 N-body simulations. The goal of the CAMELS project is to provide theory predictions for different observables as a function of cosmology and astrophysics, and it is the largest suite of cosmological (magneto-)hydrodynamic simulations designed to train machine learning algorithms. CAMELS contains thousands of different cosmological and astrophysical models by way of varying $\Omega_m$, $\sigma_8$, and four parameters controlling stellar and AGN feedback, following the evolution of more than 100 billion particles and fluid elements over a combined volume of $(400~h^{-1}{\rm Mpc})^3$. We describe the simulations in detail and characterize the large range of conditions represented in terms of the matter power spectrum, cosmic star formation rate density, galaxy stellar mass function, halo baryon fractions, and several galaxy scaling relations. We show that the IllustrisTNG and SIMBA suites produce roughly similar distributions of galaxy properties over the full parameter space but significantly different halo baryon fractions and baryonic effects on the matter power spectrum. This emphasizes the need for marginalizing over baryonic effects to extract the maximum amount of information from cosmological surveys. We illustrate the unique potential of CAMELS using several machine learning applications, including non-linear interpolation, parameter estimation, symbolic regression, data generation with Generative Adversarial Networks (GANs), dimensionality reduction, and anomaly detection., Comment: 34 pages, 19 figures, Matches published version. CAMELS webpage at https://www.camel-simulations.org

Creating accurate and low-noise covariance matrices represents a formidable challenge in modern-day cosmology. We present a formalism to compress arbitrary observables into a small number of bins by projection into a model-specific subspace that minimizes the prior-averaged log-likelihood error. The lower dimensionality leads to a dramatic reduction in covariance matrix noise, significantly reducing the number of mocks that need to be computed. Given a theory model, a set of priors, and a simple model of the covariance, our method works by using singular value decompositions to construct a basis for the observable that is close to Euclidean; by restricting to the first few basis vectors, we can capture almost all the constraining power in a lower-dimensional subspace. Unlike conventional approaches, the method can be tailored for specific analyses and captures non-linearities that are not present in the Fisher matrix, ensuring that the full likelihood can be reproduced. The procedure is validated with full-shape analyses of power spectra from BOSS DR12 mock catalogs, showing that the 96-bin power spectra can be replaced by 12 subspace coefficients without biasing the output cosmology; this allows for accurate parameter inference using only $\sim 100$ mocks. Such decompositions facilitate accurate testing of power spectrum covariances; for the largest BOSS data chunk, we find that: (a) analytic covariances provide accurate models (with or without trispectrum terms); and (b) using the sample covariance from the MultiDark-Patchy mocks incurs a $\sim 0.5\sigma$ shift in $\Omega_m$, unless the subspace projection is applied. The method is easily extended to higher order statistics; the $\sim 2000$-bin bispectrum can be compressed into only $\sim 10$ coefficients, allowing for accurate analyses using few mocks and without having to increase the bin sizes., Comment: 22 pages, 6 figures. Accepted by Phys. Rev. D

We present a new open-source code that calculates one-loop power auto- and cross-power spectra for matter fields and biased tracers in real and redshift space. These spectra incorporate all ingredients required for a direct application to data: non-linear bias, redshift-space distortions, infra-red resummation, counterterms, and the Alcock-Paczynski effect. Our code is based on the Boltzmann solver CLASS and inherits its advantage: user friendliness, ease of modification, high speed, and simple interface with other software. We present detailed descriptions of the theoretical model, the code structure, approximations, and accuracy tests. A typical end-to-end run for one cosmology takes $\sim 0.3$ seconds, which is sufficient for Markov Chain Monte Carlo parameter extraction. As an example, we apply the code to data from the Baryon Oscillation Spectroscopic Survey (BOSS) and infer cosmological parameters from the shape of the galaxy power spectrum., v1: 48 pages, 9 figures. v2: 56 pages, 14 figures, minor edits, version to appear in PRD. The code and custom-built BOSS likelihoods are available at https://github.com/Michalychforever/CLASS-PT and https://github.com/Michalychforever/lss_montepython

We introduce a physically-motivated model of the matter power spectrum, based on the halo model and perturbation theory. This model achieves 1\% accuracy on all $k-$scales between $k=0.02h\,\mathrm{Mpc}^{-1}$ to $k=1h\,\mathrm{Mpc}^{-1}$. Our key ansatz is that the number density of halos depends on the non-linear density contrast filtered on some unknown scale $R$. Using the Effective Field Theory of Large Scale Structure to evaluate the two-halo term, we obtain a model for the power spectrum with only two fitting parameters: $R$ and the effective `sound speed', which encapsulates small-scale physics. This is tested with two suites of cosmological simulations across a broad range of cosmologies and found to be highly accurate. Due to its physical motivation, the statistics can be easily extended beyond the power spectrum; we additionally derive the one-loop covariance matrices of cluster counts and their combination with the matter power spectrum. This yields a significantly better fit to simulations than previous models, and includes a new model for super-sample effects, which is rigorously tested with separate universe simulations. At low redshift, we find a significant ($\sim 10\%$) exclusion covariance from accounting for the finite size of halos which has not previously been modeled. Such power spectrum and covariance models will enable joint analysis of upcoming large-scale structure surveys, gravitational lensing surveys and cosmic microwave background maps on scales down to the non-linear scale. We provide a publicly released Python code., 40 pages, 13 figures, accepted by Phys. Rev. D. Python package and tutorial available at http://EffectiveHalos.readthedocs.io

We present cosmological constraints from a joint analysis of the pre- and post-reconstruction galaxy power spectrum multipoles from the final data release of the Baryon Oscillation Spectroscopic Survey (BOSS). Geometric constraints are obtained from the positions of BAO peaks in reconstructed spectra, analyzed in combination with the unreconstructed spectra in a full-shape (FS) likelihood using a joint covariance matrix, giving stronger parameter constraints than FS-only or BAO-only analyses. We introduce a new method for obtaining constraints from reconstructed spectra based on a correlated theoretical error, which is shown to be simple, robust, and applicable to any flavor of density-field reconstruction. Assuming $\Lambda$CDM with massive neutrinos, we analyze data from two redshift bins $z_\mathrm{eff}=0.38,0.61$ and obtain $1.6\%$ constraints on the Hubble constant $H_0$, using only a single prior on the current baryon density $\omega_b$ from Big Bang Nucleosynthesis (BBN) and no knowledge of the power spectrum slope $n_s$. This gives $H_0 = 68.6\pm1.1\,\mathrm{km\,s}^{-1}\mathrm{Mpc}^{-1}$, with the inclusion of BAO data sharpening the measurement by $40\%$, representing one of the strongest current constraints on $H_0$ independent of cosmic microwave background data. Restricting to the best-fit slope $n_s$ from Planck (but without additional priors on the spectral shape), we obtain a $1\%$ $H_0$ measurement of $67.8\pm 0.7\,\mathrm{km\,s}^{-1}\mathrm{Mpc}^{-1}$. We find strong constraints on the cosmological parameters from a joint analysis of the FS, BAO, and Planck data. This sets new bounds on the sum of neutrino masses $\sum m_\nu < 0.14\,\mathrm{eV}$ (at $95\%$ confidence) and the effective number of relativistic degrees of freedom $N_\mathrm{eff} = 2.90^{+0.15}_{-0.16}$, though contours are not appreciably narrowed by the inclusion of BAO data., Comment: 42 pages, 12 figures, accepted by JCAP, likelihoods available at https://github.com/Michalychforever/lss_montepython (minor typo corrected)

We present $\mathcal{O}(N^2)$ estimators for the small-scale power spectrum and bispectrum in cosmological simulations. In combination with traditional methods, these allow spectra to be efficiently computed across a vast range of scales, requiring orders of magnitude less computation time than Fast Fourier Transform based approaches alone. These methods are applicable to any tracer; simulation particles, halos or galaxies, and take advantage of the simple geometry of the box and periodicity to remove almost all dependence on large random particle catalogs. By working in configuration-space, both power spectra and bispectra can be computed via a weighted sum of particle pairs up to some radius, which can be reduced at larger $k$, leading to algorithms with decreasing complexity on small scales. These do not suffer from aliasing or shot-noise, allowing spectra to be computed to arbitrarily large wavenumbers. The estimators are rigorously derived and tested against simulations, and their covariances discussed. The accompanying code, HIPSTER, has been publicly released, incorporating these algorithms. Such estimators will be of great use in the analysis of large sets of high-resolution simulations., Comment: 31 pages, 6 figures, accepted by MNRAS. Code available at https://HIPSTER.readthedocs.io

The marked power spectrum is capable of placing far tighter constraints on cosmological parameters (particularly the neutrino mass) than the conventional power spectrum. What new information does it contain beyond conventional statistics? Through the development of a perturbative model, we find that the mark induces a significant coupling between non-Gaussianities, which are usually found on small scales, and large scales, leading to the additional information content. The model is derived in the context of one-loop perturbation theory and validated by comparison to $N$-body simulations across a variety of mark parameters. At moderate redshifts, including for massive neutrino cosmologies, the theory is in good agreement with the simulations. The importance of non-linear gravitational effects on the large-scale spectra complicates the modeling as there is no well-defined convergence radius of the theory at low $z$. Extension to higher perturbative order and biased tracers is possible via a similar approach, and a simple model of the latter is shown to yield promising results. The theory becomes non-perturbative at redshift zero for small smoothing scales, with important contributions from higher-order terms: these will need to be studied before the full power of this tool can be realized., Comment: 16+8 pages, 8 figures, accepted by Phys. Rev. D

We present a new class of estimators for computing small-scale power spectra and bispectra in configuration-space via weighted pair- and triple-counts, with no explicit use of Fourier transforms. Particle counts are truncated at $R_0\sim 100h^{-1}\,\mathrm{Mpc}$ via a continuous window function, which has negligible effect on the measured power spectrum multipoles at small scales. This gives a power spectrum algorithm with complexity $\mathcal{O}(NnR_0^3)$ (or $\mathcal{O}(Nn^2R_0^6)$ for the bispectrum), measuring $N$ galaxies with number density $n$. Our estimators are corrected for the survey geometry and have neither self-count contributions nor discretization artifacts, making them ideal for high-$k$ analysis. Unlike conventional Fourier transform based approaches, our algorithm becomes more efficient on small scales (since a smaller $R_0$ may be used), thus we may efficiently estimate spectra across $k$-space by coupling this method with standard techniques. We demonstrate the utility of the publicly available power spectrum algorithm by applying it to BOSS DR12 simulations to compute the high-$k$ power spectrum and its covariance. In addition, we derive a theoretical rescaled-Gaussian covariance matrix, which incorporates the survey geometry and is found to be in good agreement with that from mocks. Computing configuration- and Fourier-space statistics in the same manner allows us to consider joint analyses, which can place stronger bounds on cosmological parameters; to this end we also discuss the cross-covariance between the two-point correlation function and the small-scale power spectrum., 29 pages, 14 figures, accepted by MNRAS. Code is available at https://github.com/oliverphilcox/HIPSTER with documentation at https://HIPSTER.readthedocs.io

Constraining parameters such as the initial mass function high-mass slope and the frequency of type Ia supernovae is of critical importance in the ongoing quest to understand galactic physics and create realistic hydrodynamical simulations. In this paper, we demonstrate a method to precisely determine these using individual chemical abundances from a large set of stars, coupled with some estimate of their ages. Inference is performed via the simple chemical evolution model Chempy in a Bayesian framework, marginalizing over each star's specific interstellar medium parameters, including an element-specific `model error' parameter to account for inadequacies in our model. Hamiltonian Monte Carlo (HMC) methods are used to sample the posterior function, made possible by replacing Chempy with a trained neural network at negligible error. The approach is tested using data from both Chempy and the IllustrisTNG simulation, showing sub-percent agreement between inferred and true parameters using data from up to 1600 individual stellar abundances. For IllustrisTNG, strongest constraints are obtained from metal ratios, competitive with those from other methods including star counts. Analysis using a different set of nucleosynthetic yields shows that incorrectly assumed yield models can give non-negligible bias in the derived parameters; this is reduced by our model errors, which further show how well the yield tables match data. We also find a significant bias from analyzing only a small set of stars, as is often done in current analyses. The method can be easily applied to observational data, giving tight bounds on key galactic parameters from chemical abundances alone., 25 pages, 9 figures, accepted by ApJ. Code has been made publicly available at https://github.com/oliverphilcox/ChempyMulti

We present the one-loop perturbation theory for the power spectrum of the marked density field of matter and biased tracers in real- and redshift-space. The statistic has been shown to yield impressive constraints on cosmological parameters; to exploit this, we require an accurate and computationally inexpensive theoretical model. Comparison with $N$-body simulations demonstrates that linear theory fails on all scales, but inclusion of one-loop Effective Field Theory terms gives a substantial improvement, with $\sim 5\%$ accuracy at $z = 1$. The expansion is less convergent in redshift-space (achieving $\sim 10\%$ accuracy), but there are significant improvements for biased tracers due to the freedom in the bias coefficients. The large-scale theory contains non-negligible contributions from all perturbative orders; we suggest a reorganization of the theory that contains all terms relevant on large-scales, discussing both its explicit form at one-loop and structure at infinite-loop. This motivates a low-$k$ correction term, leading to a model that is sub-percent accurate on large scales, albeit with the inclusion of two (three) free coefficients in real- (redshift-)space. We further consider the effects of massive neutrinos, showing that beyond-EdS corrections to the perturbative kernels are negligible in practice. It remains to see whether the purported gains in cosmological parameters remain valid for biased tracers and can be captured by the theoretical model., Comment: 40 pages, 11 figures, accepted by JCAP

To fully harvest the rich library of stellar elemental abundance data available, we require reliable models that facilitate our interpretation of them. Galactic chemical evolution (GCE) models are one such set, a key part of which are the selection of chemical yields from different nucleosynthetic enrichment channels, predominantly asymptotic giant branch stars, Type Ia supernovae (SNe Ia), and core-collapse supernovae (CC-SNe). Here we present a scoring system for yield tables based on their ability to reproduce protosolar abundances within a simple parameterization of the GCE modeling software Chempy, which marginalizes over galactic parameters describing simple stellar populations (SSPs) and interstellar medium physics. Two statistical scoring methods are presented, based on Bayesian evidence and leave-one-out cross- validation, and are applied to five CC-SN tables, for (a) all mutually available elements and (b) a subset of the nine most abundant elements. We find that the yields of Prantzos et al. (P18, including stellar rotation) and Chieffi & Limongi (C04) best reproduce protosolar abundances for the two cases, respectively. The inferred best-fit SSP parameters for case (b) are {α }IMF}=-{2.45}-0.11+0.15 for the initial mass function high-mass slope and {N}Ia}={1.29}-0.31+0.45× {10}-3 {M}☉ -1 for the SN Ia normalization, which are broadly consistent across tested yield tables. Additionally, we demonstrate how Chempy can be used to dramatically improve elemental abundance predictions of hydrodynamical simulations by plugging tailored best-fit SSP parameters into a Milky Way analog from Gutcke & Springel. Our code, including a comprehensive tutorial, is freely available and can additionally provide SSP enrichment tables for any combination of parameters and yield tables.

Searches for inflationary gravitational wave signals in the CMB B-mode polarisation are expected to reach unprecedented power over the next decade. A major difficulty in these ongoing searches is that galactic foregrounds such as dust can easily mimic inflationary signals. Though typically foregrounds are separated from primordial signals using the foregrounds' different frequency dependence, in this paper we investigate instead the extent to which the galactic dust B-modes' statistical anisotropy can be used to distinguish them from inflationary B-modes, building on the work of Kamionkowski and Kovetz (2014). In our work, we extend existing anisotropy estimators and apply them to simulations of polarised dust to forecast their performance for future experiments. Considering the application of this method as a null-test for dust contamination to CMB-S4, we find that we can detect residual dust levels corresponding to $r\sim0.001$ at $2\sigma$, which implies that statistical anisotropy estimators will be a powerful diagnostic for foreground residuals (though our results show some dependence on the dust simulation used). Finally, considering applications beyond a simple null test, we demonstrate how anisotropy statistics can be used to construct an estimate of the dust B-mode map, which could potentially be used to clean the B-mode sky., Comment: 20 pages, 13 figures, accepted by MNRAS