Using X-Ray Morphological Parameters to Strengthen Galaxy Cluster Mass Estimates via Machine Learning

Daisuke Nagai; Dominique Eckert; John A. ZuHone; Klaus Dolag; Lorenzo Lovisari; Michelle Ntampaka; Sheridan B. Green

REVIEW 2 major objections 6 minor 1 cited by

Using X-Ray Morphological Parameters to Strengthen Galaxy Cluster Mass Estimates via Machine Learning

T0 review · 2 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A random forest that reads a galaxy cluster's X-ray shape alongside its luminosity cuts mass-estimate scatter by 20 percent relative to luminosity alone.

desk verdict A solid ML proof-of-concept on cluster masses; the 20% scatter reduction is plausible, but the exact-R500c aperture assumption makes the survey-ready claim optimistic rather than proven. read the letter →

arxiv 1908.02765 v2 pith:IW4K3AUD submitted 2019-08-07 astro-ph.CO

Sheridan B. Green , Michelle Ntampaka , Daisuke Nagai , Lorenzo Lovisari , Klaus Dolag , Dominique Eckert , John A. ZuHone This is my paper · ORCID

classification astro-ph.CO

keywords galaxyclustersX-raymorphologyrandomforestmachinelearningclustermassestimationcore-excisedluminositydynamicalstatesurveys

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper argues that the dynamical state of a galaxy cluster, read from the shape of its X-ray image, carries information about the cluster's total mass that luminosity alone misses. Training a random forest on core-excised X-ray luminosity plus morphological parameters -- concentration, smoothness, asymmetry, power ratios, centroid shift, ellipticity, and $M_{20}$ -- reduces scatter in predicted masses by about 20 percent relative to a standard luminosity-only regression. The improvement appears in both an idealized high-resolution, long-exposure mock series and a realistic short-exposure survey mock series with background, each reaching $1\sigma$ scatter of about $0.066$ dex (16 percent) with negligible bias. If this transfers to real observations, upcoming X-ray surveys can estimate masses of low-photon clusters more accurately without requiring expensive spectra.

What carries the argument

The load-bearing mechanism is the random forest regressor -- an ensemble of decision trees whose predictions are averaged -- fed with a feature set consisting of core-excised, redshift-scaled X-ray luminosity plus morphological parameters. The morphological parameters translate a cluster's dynamical state into image-derived numbers: surface brightness concentration, smoothness, asymmetry, power ratios, centroid shift, ellipticity, and $M_{20}$. The forest learns nonlinear combinations of these features; smoothness, asymmetry, and concentration carry most of the mass information after luminosity, while the remaining parameters as a group still supply about a third of the total improvement.

What would settle it

Apply the trained random forest to a real X-ray cluster sample with independent masses from weak lensing or $Y_X$; if the scatter relative to those masses is not about 20 percent below a luminosity-only regression, or if recomputing the features with observationally estimated $R_{500c}$ and PSF deconvolution erases the gap, the central claim is falsified.

Watch

Extended reading notes

Core claim

The central claim is that mass-encoding dynamical state information, quantified by X-ray morphological parameters, remains present even in low-photon survey observations, and that a random forest can extract it. Trained on mock observations of 2,041 simulated clusters and tested on a held-out 20 percent, the random forest predicts $\log(M_{500c})$ with $1\sigma$ intrinsic scatter $\delta = 0.066$ dex in both the idealized and realistic mock series. That is a 20 percent reduction in scatter relative to the standard core-excised luminosity relation, whose scatter is $\delta = 0.081$ dex, and the predicted masses show negligible bias. Linear regressions using the same morphological features improve only marginally over luminosity alone, indicating that the gain is a nonlinear effect captured by the forest.

Load-bearing premise

The result is demonstrated only on simulated clusters with known masses; the 20 percent improvement transfers to real surveys only if the mapping from X-ray morphology and luminosity to mass learned from those simulations is faithful to actual clusters, given that the features use the true cluster radius and no PSF deconvolution.

Editorial extensions

If this is right

Upcoming wide-area X-ray surveys can assign cluster masses with roughly 16 percent scatter for low-photon systems, without measuring gas temperature.
The method uses the full photon distribution rather than only core-excised counts, so the gain should persist near the detection threshold.
Smoothness, asymmetry, and concentration are the highest-value features, so future surveys and pipelines can prioritize measuring them reliably.
Including the additional morphological parameters beyond those three still matters, contributing about a third of the total scatter reduction.
The systematic underprediction of high-mass clusters should be reduced by training on a sample with a flat mass function across the full mass range of interest.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

As an editorial extension, the same morphology features could be combined with SZ or optical richness proxies, since dynamical state is known to affect scatter in those mass-observable relations too.
A direct testable extension would retrain the model on mocks where $R_{500c}$ is estimated rather than taken from the simulation and where PSF smearing is deconvolved; if the 20 percent gain shrinks, the realistic survey gain is smaller than reported.
The feature-importance ranking suggests a simpler diagnostic, such as smoothness alone, might recover much of the improvement on real data, which could be checked with a small pilot sample.
The apparent universality of the morphology-mass mapping across redshifts $0.1 \le z \le 0.29$ is an extrapolation beyond the trained range and would need validation at higher redshift.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Main take: this is a well-executed ML proof-of-concept, not a breakthrough. What's new is the specific combination—random forest on core-excised luminosity plus a set of X-ray morphological parameters—evaluated on both idealized Chandra and realistic eROSITA mocks. The experimental design is one of the better ones in this literature: train/test split by unique cluster, 10-fold cross-validation, a held-out test set, and a check with a reduced feature set. The headline result, a 20% scatter reduction from 0.081 to 0.066 dex, is plausible, and the feature-importance finding that smoothness, asymmetry, and concentration carry most of the mass information is a useful output. The authors also report the high-mass underprediction honestly and note that linear models barely improve over the luminosity baseline, which strengthens the claim that the gain is genuinely nonlinear.

The soft spots are real but not disqualifying. The 0.066 dex figure has no uncertainty estimate or significance test; on a 426-cluster test set, a 0.015 dex improvement should come with bootstrap error bars. More substantively, all features are computed inside the exact simulation R500c, and R500c is defined by the target M500c. That means the aperture itself leaks target information into the features. Notably, the M-Lex,z baseline uses the same true aperture, so the in-mock comparison is fair; the issue is only about generalization to real data. The authors acknowledge this in Section 3 and call their results optimistic, but they never quantify how much of the improvement survives when R500c must be estimated from noisy data. That is the key question for eROSITA, and it is left open. The single-simulation-family issue is another limitation, though common in this subfield.

Who this is for: cluster cosmologists working on mass proxies, or anyone testing ML on mock observations. It deserves peer review: the question is well-posed, the methodology is careful, and the limitations are explicit. I would not desk-reject. I'd ask the referee to request either a noisy-aperture rerun or an explicit statement that the survey-ready claim depends on that missing test. Bottom line: worth reading and citing as a proof-of-concept, but the survey forecast should not be quoted without the caveat.

Referee Report

2 major / 6 minor

Summary. This paper trains a random forest regressor to estimate galaxy cluster masses (M500c) from X-ray observables, using core-excised luminosity and a set of surface-brightness morphological parameters (concentration, centroid shift, power ratios, ellipticity, asymmetry, smoothness, and M20). The training data are 2,041 mock Chandra-like and eROSITA-like observations of 984 unique clusters drawn from Magneticum simulations; features are computed within the true R500c aperture. The model is validated with a cluster-split train/test partition and 10-fold cross-validation, and compared against standard mass-luminosity regression and several linear models. The authors report a 20% reduction in the 1σ scatter of mass residuals (0.066 dex for the random forest versus 0.081 dex for M-Lex,z) on the test set, for both the idealized Chandra and realistic eROSITA mock series, with feature importance dominated by core-excised luminosity, smoothness, asymmetry, and concentration. The paper concludes that morphological parameters can improve cluster mass estimates in upcoming surveys such as eROSITA.

Significance. If the reported improvement is robust, this is a valuable contribution to cluster cosmology: a photometric-only mass proxy that beats the standard M-Lex,z relation on mock data, with the same performance in a short-exposure, low-photon eROSITA-like setting as in idealized Chandra-like observations. The study has several methodological strengths: the train/test split is performed by unique cluster identity rather than by individual observations, which prevents cross-contamination; the 10-fold cross-validation folds are also cluster-split; two mock series bracket the range from idealized to realistic X-ray conditions; and the comparison across linear, regularized, and nonlinear regressors gives a clear picture of where the gain comes from. These design choices make the internal result credible. The main significance hinges on whether the performance gain survives realistic aperture estimation, since the features are computed using the exact simulation-defined R500c, whose definition is tied to the very mass being predicted.

major comments (2)

[Section 3 and Section 6]
[Section 5, Table 2, Figure 5]

minor comments (6)

[Abstract and Section 4.1]
[Section 4.2]
[Figure 4 caption]
[Table 1]
[Section 3]
[Section 5]

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the 20% scatter reduction is a held-out supervised-learning benchmark; the exact-R500c aperture is an acknowledged idealization, not a re-fit.

full rationale

The central claim is an empirical comparison of regression models on a test set that was never used for training or hyperparameter selection, with the train/test split performed by unique cluster ID so that multiple redshift observations of the same cluster do not leak across the split. No fitted parameter is presented as a prediction: the random forest is trained and cross-validated on training clusters and then evaluated on held-out clusters, and the M−Lex,z baseline is also fit to the same training sample. The morphological features are standard literature definitions, and the cited prior ML work, including the authors' own Ntampaka et al. (2019b), is contextual rather than load-bearing: no uniqueness theorem, fitted ansatz, or calibration constant is imported from those papers to force the present result. The only self-referential aspect is that both training and test clusters come from the same Magneticum/PHOX simulation family, which is a domain-transfer limitation rather than an internal circularity. The paper explicitly flags its main idealization in Section 3: by using the exact R500c for the aperture and, for the Chandra series, no PSF deconvolution, the results are 'optimistic estimates'; it also notes in the conclusion that scatter in R500c measurements will propagate into the morphological parameters. This is a candid limitation on transfer to real surveys, not a definition that equates the output with the input. The 20% improvement over M−Lex,z is computed with the same idealized aperture for both models, so the relative comparison is internally consistent. No circular step is exhibited by the paper's equations or by its self-citations, so the appropriate finding is no significant circularity.

Assumptions & free parameters 1 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new physical entities and derives nothing from first principles. It is a supervised regression benchmark built on the Magneticum simulations and PHOX mock observations. The central claim therefore rests on the fidelity of the simulation pipeline and on several idealizations (true R500c apertures, self-similar redshift scaling, SUBFIND masses). The only fitted numbers are the random forest hyperparameters selected by cross-validation. The main risk is not internal circularity but domain transfer from simulations to real observations.

free parameters (1)

Random forest hyperparameters = not specified; selected via grid-search cross-validation
The CV-tuned model's hyperparameters (tree count, max features, depth, split and leaf sample counts, bootstrap) are fitted to the training data and are not reported. The default RF performs nearly as well (0.067-0.070 dex), so the headline result is only mildly sensitive to them.

assumptions (5)

domain assumption Magneticum/PHOX mocks faithfully reproduce X-ray cluster emission, morphology, and instrument response for Chandra and eROSITA.
All training features and test images come from this pipeline. If the simulated morphology-mass relation is unrealistic, the result is a simulation-only benchmark rather than a statement about real clusters.
domain assumption Self-similar redshift evolution of the luminosity-mass relation, Lex,z = LX,ex E(z)^-7/3 (Eq. 1).
Used to combine clusters at z = 0.1-0.29 into one sample; no evolution beyond self-similar is modeled.
domain assumption Apertures use the true R500c and 0.15 R500c for core excision, known from the simulation.
Real surveys must estimate R500c, which adds scatter. The authors explicitly call the results optimistic estimates because of this idealization.
domain assumption SUBFIND M500c masses are accurate enough to serve as regression targets.
Training labels come from SUBFIND on Magneticum. The paper cites the Knebe et al. (2011) halo finder comparison suggesting few-percent accuracy for host halo masses.
standard math Random forest generalizes from training to held-out clusters within the same simulation.
Standard ML assumption validated by the test split. Random forests cannot extrapolate outside the training mass range, and the paper observes high-mass underprediction consistent with this.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Using X-Ray Morphological Parameters to Strengthen Galaxy Cluster Mass Estimates via Machine Learning." pith.science (2026). https://pith.science/paper/IW4K3AUD

@misc{pith2026190802765,
  author       = {Pith},
  title        = {Pith review of: Using X-Ray Morphological Parameters to Strengthen Galaxy Cluster Mass Estimates via Machine Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IW4K3AUD}},
  note         = {Machine review of arXiv:1908.02765}
}

read the original abstract

We present a machine learning approach for estimating galaxy cluster masses, trained using both Chandra and eROSITA mock X-ray observations of 2,041 clusters from the Magneticum simulations. We train a random forest regressor, an ensemble learning method based on decision tree regression, to predict cluster masses using an input feature set. The feature set uses core-excised X-ray luminosity and a variety of morphological parameters, including surface brightness concentration, smoothness, asymmetry, power ratios, and ellipticity. The regressor is cross-validated and calibrated on a training sample of 1,615 clusters (80% of sample), and then results are reported as applied to a test sample of 426 clusters (20% of sample). This procedure is performed for two different mock observation series in an effort to bracket the potential enhancement in mass predictions that can be made possible by including dynamical state information. The first series is computed from idealized Chandra-like mock cluster observations, with high spatial resolution, long exposure time (1 Ms), and the absence of background. The second series is computed from realistic-condition eROSITA mocks with lower spatial resolution, short exposures (2 ks), instrument effects, and background photons modeled. We report a 20% reduction in the mass estimation scatter when either series is used in our random forest model compared to a standard regression model that only employs core-excised luminosity. The morphological parameters that hold the highest feature importance are smoothness, asymmetry, and surface brightness concentration. Hence, these parameters, which encode the dynamical state of the cluster, can be used to make more accurate predictions of cluster masses in upcoming surveys, offering a crucial step forward for cosmological analyses.

Figures

Figures reproduced from arXiv: 1908.02765 by the authors.

**Figure 1.** Mass function of the cluster sample used in this work. The sample is flat in the range 1013.5 ≤ M500c/(h −1 M ) ≤ 1014.2 and begins decaying in cluster counts for 1014.2 ≤ M500c/(h −1 M ) ≤ 1014.8 . This sample consists of a total of 2,041 clusters. This uniform distribution in log(M500c), our predicted quantity, enables the regression model optimization to equally weight a broad range of cluster masses. sis). Box2b… view at source ↗

**Figure 2.** Distributions of the surface brightness concentration c, asymmetry A, and smoothness S computed from the “idealized Chandra” and “realistic eROSITA” mock cluster observation series. Several other morphological parameters are also employed in the analysis (see Sec. 3), but we find that c, A, and S are most important for strengthening the cluster mass model. and high-photon count observations. The latter series is in… view at source ↗

**Figure 3.** Sample Chandra-like cluster images illustrating morphological parameter differences. Each image is centered on the cluster X-ray peak and is cropped a side length of 2R500c. All clusters shown have 1014.3 ≤ M500c/(h −1 M ) ≤ 1014.6 at z = 0.1. Top: highly concentrated cluster on the left (c = 0.37) and weakly concentrated cluster on the right (c = 0.04). Middle: Asymmetric cluster on the left (A = 1.49) and symmetr… view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: Predicted mass as a function of true mass. Predictions are made using the cross-validation-tuned RF models, which are separately trained using each of the two morphological parameter series. The distributions both have low intrinsic scatter δ (0.066 dex) and a negligi…

**Figure 5.** Figure 5: PDF of mass residuals for the cross-validation-tuned RF models of both the “idealized Chandra” and “realistic eROSITA” observation series. For comparison, we plot the PDF of mass residuals for the mass-luminosity relationship, with Lex,z computed using core-excised lu…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

Tighter Dark Matter Constraints from the Projected Mass Method: A Neural Network Enhanced Method for Galaxy Groups and Clusters
astro-ph.GA 2026-07 conditional novelty 5.0 of 10

A simulation-trained neural network removes most of the systematic overestimate of the Projected Mass Estimator, yielding Milky Way, M81, and NGC 5128 halo masses in line with the literature.

Reference graph

Works this paper leans on

74 extracted references · 42 canonical work pages · cited by 1 Pith paper

[1]

G., Tanvir, N

Abraham, R. G., Tanvir, N. R., Santiago, B. X., et al. 1996, MNRAS, 279, L47

work page 1996
[2]

E., Springel, V ., White, S

Angulo, R. E., Springel, V ., White, S. D. M., et al. 2012, Monthly Notices of the Royal Astronomical Society, 426, 2046

work page 2012
[3]

E., Mantz, A., Allen, S

Applegate, D. E., Mantz, A., Allen, S. W., et al. 2016, MNRAS, 457, 1522

work page 2016
[4]

J., Kay, S

Armitage, T. J., Kay, S. T., & Barnes, D. J. 2019, MNRAS, 484, 1526

work page 2019
[5]

Arnaud, K. A. 1996, in Astronomical Society of the Paciﬁc Conference Series, V ol. 101, Astronomical Data Analysis Software and Systems V , ed. G. H. Jacoby & J. Barnes, 17 Bifﬁ, V ., Dolag, K., & B¨ohringer, H. 2013, MNRAS, 428, 1395 Bifﬁ, V ., Dolag, K., B¨ohringer, H., & Lemson, G. 2012, MNRAS, 420, 3545 Bifﬁ, V ., Borgani, S., Murante, G., et al. 2016...

work page 1996
[6]

Bocquet, S., Saro, A., Dolag, K., & Mohr, J. J. 2016, MNRAS, 456, 2361

2016
[7]

Bolliet, B., Comis, B., Komatsu, E., & Mac´ıas-P´erez, J. F. 2018, MNRAS, 477, 4957

work page 2018
[8]

H., Mohammed, I., & Lovisari, L

Borm, K., Reiprich, T. H., Mohammed, I., & Lovisari, L. 2014, A&A, 567, A65

work page 2014

Show all 74 references

[9]

2001, Mach

Breiman, L. 2001, Mach. Learn., 45, 5

2001
[10]

A., & Tsai, J

Buote, D. A., & Tsai, J. C. 1995, ApJ, 452, 522

1995
[11]

F., & Berlind, A

Calderon, V . F., & Berlind, A. A. 2019, arXiv e-prints, arXiv:1902.02680

2019 arXiv
[12]

E., Ridl, J., et al

Clerc, N., Ramos-Ceja, M. E., Ridl, J., et al. 2018, A&A, 617, A92

2018
[13]

D., & Battaglia, N

Cohn, J. D., & Battaglia, N. 2019, arXiv e-prints, arXiv:1905.09920

2019 arXiv
[14]

Conselice, C. J. 2003, ApJS, 147, 1 de Haan, T., Benson, B. A., Bleem, L. E., et al. 2016, ApJ, 832, 95 DES Collaboration, Abbott, T. M. C., Abdalla, F. B., et al. 2017, ArXiv e-prints, arXiv:1708.01530

2003 arXiv
[15]

P., Bocquet, S., Schrabback, T., et al

Dietrich, J. P., Bocquet, S., Schrabback, T., et al. 2019, MNRAS, 483, 2871

2019
[16]

2009, MNRAS, 399, 497

Dolag, K., Borgani, S., Murante, G., & Springel, V . 2009, MNRAS, 399, 497

2009
[17]

M., Beck, A

Dolag, K., Gaensler, B. M., Beck, A. M., & Beck, M. C. 2015, MNRAS, 451, 4277

2015
[18]

2016, MNRAS, 463, 1797

Dolag, K., Komatsu, E., & Sunyaev, R. 2016, MNRAS, 463, 1797

2016
[19]

2011, A&A, 526, A79

Eckert, D., Molendi, S., & Paltani, S. 2011, A&A, 526, A79

2011
[20]

2019, A&A, 621, A40

Eckert, D., Ghirardini, V ., Ettori, S., et al. 2019, A&A, 621, A40

2019
[21]

2019, A&A, 621, A39

Ettori, S., Ghirardini, V ., Eckert, D., et al. 2019, A&A, 621, A39

2019
[22]

Freund, Y ., & Schapire, R. E. 1996, in Proceedings of the 13th International Conference on Machine Learning (Morgan Kaufmann), 148–156

1996
[23]

Friedman, J. H. 2001, Ann. Statist., 29, 1189 G´eron, A. 2017, Hands-On Machine Learning with Scikit-Learn and TensorFlow: Concepts, Tools, and Techniques to Build Intelligent Systems, 1st edn. (O’Reilly Media, Inc.)

2001
[24]

2019, A&A, 621, A41

Ghirardini, V ., Eckert, D., Ettori, S., et al. 2019, A&A, 621, A41

2019
[25]

A., Puchwein, E., Shen, S., & Sijacki, D

Henden, N. A., Puchwein, E., Shen, S., & Sijacki, D. 2018, MNRAS, 479, 5385 12 G REEN ET AL

2018
[26]

2017, MNRAS, 465, 3361

Schaye, J. 2017, MNRAS, 465, 3361

2017
[27]

L., et al

Hildebrandt, H., K¨ohlinger, F., van den Busch, J. L., et al. 2018, arXiv e-prints, arXiv:1812.06076

2018 arXiv
[28]

M., Ntampaka, M., et al

Ho, M., Rau, M. M., Ntampaka, M., et al. 2019, arXiv e-prints, arXiv:1902.05950

2019 arXiv
[29]

2015, MNRAS, 449, 685

Hoekstra, H., Herbonnet, R., Muzzin, A., et al. 2015, MNRAS, 449, 685

2015
[30]

E., & Kennard, R

Hoerl, A. E., & Kennard, R. W. 1970, Technometrics, 12, 55

1970
[31]

Reiprich, T. H. 2015, MNRAS, 448, 814

2015
[32]

2019, arXiv e-prints, arXiv:1906.09262

Joudaki, S., Hildebrandt, H., Traykova, D., et al. 2019, arXiv e-prints, arXiv:1906.09262

2019 arXiv
[33]

R., Muldrew, S

Knebe, A., Knollmann, S. R., Muldrew, S. I., et al. 2011, MNRAS, 415, 2293

2011
[34]

M., Dunkley, J., et al

Komatsu, E., Smith, K. M., Dunkley, J., et al. 2011, ApJS, 192, 18

2011
[35]

V ., & Borgani, S

Kravtsov, A. V ., & Borgani, S. 2012, Annu. Rev. Astron. Astrophys., 50, 353

2012
[36]

V ., Vikhlinin, A., & Nagai, D

Kravtsov, A. V ., Vikhlinin, A., & Nagai, D. 2006, ApJ, 650, 128

2006
[37]

T., Kravtsov, A

Lau, E. T., Kravtsov, A. V ., & Nagai, D. 2009, ApJ, 705, 1129

2009
[38]

T., Nagai, D., & Nelson, K

Lau, E. T., Nagai, D., & Nelson, K. 2013, ApJ, 777, 151

2013
[39]

M., Primack, J., & Madau, P

Lotz, J. M., Primack, J., & Madau, P. 2004, AJ, 128, 163

2004
[40]

R., Jones, C., et al

Lovisari, L., Forman, W. R., Jones, C., et al. 2017, ApJ, 846, 51

2017
[41]

2019, arXiv e-prints, arXiv:1907.07870

Makiya, R., Hikage, C., & Komatsu, E. 2019, arXiv e-prints, arXiv:1907.07870

2019 arXiv
[42]

2019, arXiv e-prints, arXiv:1907.01560

Man, Z.-y., Peng, Y .-j., Shi, J.-j., et al. 2019, arXiv e-prints, arXiv:1907.01560

2019 arXiv
[43]

B., Allen, S

Mantz, A. B., Allen, S. W., Morris, R. G., & von der Linden, A. 2018, MNRAS, 473, 3072

2018
[44]

P., Smith, G

Marrone, D. P., Smith, G. P., Okabe, N., et al. 2012, ApJ, 754, 119

2012
[45]

Maughan, B. J. 2007, ApJ, 668, 772

2007
[46]

2012, arXiv e-prints, arXiv:1209.3114

Merloni, A., Predehl, P., Becker, W., et al. 2012, arXiv e-prints, arXiv:1209.3114

2012 arXiv
[47]

J., Fabricant, D

Mohr, J. J., Fabricant, D. G., & Geller, M. J. 1993, ApJ, 413, 492

1993
[48]

Nagai, D., Vikhlinin, A., & Kravtsov, A. V . 2007, ApJ, 655, 98

2007
[49]

2019a, arXiv e-prints, arXiv:1906.07729

Ntampaka, M., Rines, K., & Trac, H. 2019a, arXiv e-prints, arXiv:1906.07729

1906 arXiv
[50]

J., et al

Ntampaka, M., Trac, H., Sutherland, D. J., et al. 2015, ApJ, 803, 50 —. 2016, ApJ, 831, 135

2015
[51]

Pedregosa, F., Varoquaux, G., Gramfort, A., et al. 2011, J. Mach. Learn. Res., 12, 2825

2011
[52]

Pillepich, A., Porciani, C., & Reiprich, T. H. 2012, MNRAS, 422, 44

2012
[53]

H., Porciani, C., Borm, K., & Merloni, A

Pillepich, A., Reiprich, T. H., Porciani, C., Borm, K., & Merloni, A. 2018, MNRAS, 481, 613 Planck Collaboration, Aghanim, N., Arnaud, M., et al. 2011, A&A, 536, A9 Planck Collaboration, Ade, P. A. R., Aghanim, N., et al. 2016, A&A, 594, A24

2018
[54]

W., Arnaud, M., Biviano, A., et al

Pratt, G. W., Arnaud, M., Biviano, A., et al. 2019, SSRv, 215, 25

2019
[55]

W., Croston, J

Pratt, G. W., Croston, J. H., Arnaud, M., & B¨ohringer, H. 2009, A&A, 498, 361

2009
[56]

Quinlan, J. R. 1986, Mach. Learn., 1, 81

1986
[57]

2017, A&C, 20, 52

Ragagnin, A., Dolag, K., Bifﬁ, V ., et al. 2017, A&C, 20, 52

2017
[58]

2019, ApJ, 872, 170

Raghunathan, S., Patil, S., Baxter, E., et al. 2019, ApJ, 872, 170

2019
[59]

2013, AstRv, 8, 40

Rasia, E., Meneghetti, M., & Ettori, S. 2013, AstRv, 8, 40

2013
[60]

2006, MNRAS, 369, 2013

Rasia, E., Ettori, S., Moscardini, L., et al. 2006, MNRAS, 369, 2013

2006
[61]

T., Borgani, S., et al

Rasia, E., Lau, E. T., Borgani, S., et al. 2014, ApJ, 791, 96

2014
[62]

2017, MNRAS, 464, 3742

Remus, R.-S., Dolag, K., Naab, T., et al. 2017, MNRAS, 464, 3742

2017
[63]

S., Rosati, P., Tozzi, P., et al

Santos, J. S., Rosati, P., Tozzi, P., et al. 2008, A&A, 483, 35

2008
[64]

Shi, X., Komatsu, E., Nagai, D., & Lau, E. T. 2016, MNRAS, 455, 2936

2016
[65]

2015, MNRAS, 448, 1020

Shi, X., Komatsu, E., Nelson, K., & Nagai, D. 2015, MNRAS, 448, 1020

2015
[66]

Shirasaki, M., Nagai, D., & Lau, E. T. 2016, Monthly Notices of the Royal Astronomical Society, 460, 3913

2016
[67]

Springel, V ., White, S. D. M., Tormen, G., & Kauffmann, G. 2001, MNRAS, 328, 726

2001
[68]

K., Dolag, K., Comerford, J

Steinborn, L. K., Dolag, K., Comerford, J. M., et al. 2016, MNRAS, 458, 1013

2016
[69]

2015, MNRAS, 448, 1504

Remus, R.-S. 2015, MNRAS, 448, 1504

2015
[70]

A., & Zeldovich, Y

Sunyaev, R. A., & Zeldovich, Y . B. 1972, Comments on Astrophysics and Space Physics, 4, 173

1972
[71]

F., Remus, R.-S., Dolag, K., et al

Teklu, A. F., Remus, R.-S., Dolag, K., et al. 2015, ApJ, 812, 29

2015
[72]

Tibshirani, R. 1996, J. Royal Stat. Soc. B, 58, 267

1996
[73]

A., V oit, G

Ventimiglia, D. A., V oit, G. M., Donahue, M., & Ameglio, S. 2008, ApJ, 685, 118 von der Linden, A., Allen, M. T., Applegate, D. E., et al. 2014, MNRAS, 439, 2

2008
[74]

A., et al

Zhang, Y .-Y ., Andernach, H., Caretta, C. A., et al. 2011, A&A, 526, A105 Zubeldia, ´I., & Challinor, A. 2019, arXiv e-prints, arXiv:1904.07887

2011 arXiv

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

G., Tanvir, N

Abraham, R. G., Tanvir, N. R., Santiago, B. X., et al. 1996, MNRAS, 279, L47

work page 1996

[2] [2]

E., Springel, V ., White, S

Angulo, R. E., Springel, V ., White, S. D. M., et al. 2012, Monthly Notices of the Royal Astronomical Society, 426, 2046

work page 2012

[3] [3]

E., Mantz, A., Allen, S

Applegate, D. E., Mantz, A., Allen, S. W., et al. 2016, MNRAS, 457, 1522

work page 2016

[4] [4]

J., Kay, S

Armitage, T. J., Kay, S. T., & Barnes, D. J. 2019, MNRAS, 484, 1526

work page 2019

[5] [5]

Arnaud, K. A. 1996, in Astronomical Society of the Paciﬁc Conference Series, V ol. 101, Astronomical Data Analysis Software and Systems V , ed. G. H. Jacoby & J. Barnes, 17 Bifﬁ, V ., Dolag, K., & B¨ohringer, H. 2013, MNRAS, 428, 1395 Bifﬁ, V ., Dolag, K., B¨ohringer, H., & Lemson, G. 2012, MNRAS, 420, 3545 Bifﬁ, V ., Borgani, S., Murante, G., et al. 2016...

work page 1996

[6] [6]

Bocquet, S., Saro, A., Dolag, K., & Mohr, J. J. 2016, MNRAS, 456, 2361

2016

[7] [7]

Bolliet, B., Comis, B., Komatsu, E., & Mac´ıas-P´erez, J. F. 2018, MNRAS, 477, 4957

work page 2018

[8] [8]

H., Mohammed, I., & Lovisari, L

Borm, K., Reiprich, T. H., Mohammed, I., & Lovisari, L. 2014, A&A, 567, A65

work page 2014