Pith. sign in

REVIEW 2 major objections 4 minor 58 references

The IMF package: a toolkit implementing mass functions and statistical tools to analyze them

T0 review · 2 major / 4 minor · reviewed 2026-08-01 · deepseek-v4-flash

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

Pith's one-line read The paper presents the imf Python package, which implements common and variant forms of mass functions — especially the stellar IMF and its pre- and protostellar equivalents — as probability distributions, enabling sampling and integration

desk verdict Useful, well-integrated IMF/PMF/CMF sampling toolkit, but the KK09 CDF in Eq (6) is mathematically wrong and needs fixing before that distribution can be trusted. read the letter →

arxiv 2607.26007 v1 pith:RWI22HUN submitted 2026-07-28 astro-ph.IM astro-ph.GAastro-ph.HEastro-ph.SR

Theo Richardson , Adam Ginsburg , Sergey Koposov This is my paper
classification astro-ph.IMastro-ph.GAastro-ph.HEastro-ph.SR
keywords initialmassfunctionPythonsoftwaresamplingcoreprotostellarstellarpopulationsinterpolationoptimal
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

The paper tries to establish that the imf package provides a working, accessible toolkit for astrophysical mass functions, filling a gap left by larger synthesis codes and single-purpose implementations. It models mass functions as probability distributions built on standard scientific Python statistical machinery, so users can sample, integrate, and characterize populations with minimal effort. The package covers canonical forms (Salpeter, Kroupa, Chabrier) as well as harder-to-implement ones (error-convolved power laws, turbulent-IMF theories, protostellar and core mass functions). If the claims hold, researchers gain a unified, flexible tool for turning mass-function models into synthetic stellar populations and for testing star-formation theories against observations.

What carries the argument

The central object is the MassFunction class, built on a Distribution layer that wraps or reimplements statistical distributions. Complex or non-analytic functions are evaluated on a grid and interpolated with PCHIP (piecewise cubic Hermite interpolating polynomial) using a default of 200 points. Random sampling works through the percent-point function (inverse CDF) of the underlying distribution; optimal sampling solves a root-finding equation for the most massive member and then iteratively integrates one-member mass intervals. CompositeDistribution joins component distributions for broken-power-law and Chabrier forms.

What would settle it

Pick a stress case — e.g., the Koen–Kondlo error-convolved power law with α=2.35 and σ=2 over (0.03, 120) M⊙ — evaluate the PDF and CDF on a dense grid via direct numeric integration, and compare with the package's interpolated values at the same masses; a deviation above ~0.1% anywhere in the range would refute the fidelity claim underpinning the package's sampling and integration.

Watch

Extended reading notes

Core claim

The paper claims that the imf package provides a unified, statistically grounded implementation of common mass functions — Salpeter, Kroupa, Chabrier, Schechter, the Koen–Kondlo error-convolved power law, and the Padoan–Nordlund turbulent IMF — and extends the same machinery to protostellar mass functions (PMFs) and core mass functions (CMFs) from several formation models. It further claims that both random sampling (via inverse-CDF draws) and deterministic optimal sampling (via iterative root-finding on one-member integrals) are implemented faithfully, with interpolation used only where analytic forms are intractable. The package also includes utilities for star-cluster multiplicity, integr

Load-bearing premise

The load-bearing numerical premise is that 200-point PCHIP interpolation keeps every interpolated PDF/CDF within about 0.1% of its true value; if that fails for a particular function, all sampling and derived statistics inherit the bias.

Editorial extensions

If this is right

  • Researchers can generate synthetic stellar populations with reproducible random or optimal sampling from any included IMF form, with configurable stop criteria controlling how the mass budget is filled.
  • Pre- and protostellar models — PMFs from the McKee–Offner formalism and CMFs from Padoan–Nordlund and Hennebelle–Chabrier — become directly testable against observations without requiring bespoke code.
  • Integrated galaxy IMFs can be assembled by sampling cluster masses from a Schechter function and then stars from an IMF, all within one package, enabling population-level comparisons.
  • The package's luminosity and Lyman-continuum helpers connect a sampled cluster to observable integrated light, providing inputs for stellar population synthesis codes.
  • Performance measurements indicate that random sampling of a 10^4 M⊙ cluster is sub-second for all tested IMFs, with memory scaling linearly, supporting repeated Monte Carlo experiments.

Reading between the lines

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

  • The paper's ~0.1% accuracy claim for PCHIP interpolation is asserted without a presented validation; a targeted error curve would tell users when to increase the default 200-point grid.
  • Because interpolated functions default to finite mass bounds, any user who requests non-finite limits gets silently truncated distributions, which could bias the high-mass tails of cluster statistics.
  • The package's PMF/CMF implementations put competing star-formation theories on equal computational footing, so a natural next step is to run the same cluster-sampling experiments across PN, HC, and M10/O11 forms to see which matches observed core and protostar counts.
  • The system-versus-stellar IMF distinction, including multiplicity assignment, is built in; users working with single-star IMFs need to be aware which population they are sampling.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 4 minor

Summary. The paper presents `imf`, a Python package implementing a variety of mass functions (Salpeter, Kroupa, Chabrier, Schechter, KK09 error-convolved power law, Padoan turbulent IMF) as SciPy-compatible distributions, with random and optimal sampling, star-cluster generation, integrated-galaxy IMF construction, CMF/PMF models, luminosity calculations, and visualization. The authors describe the mathematical forms, the interpolation strategy used for expensive functions, the sampling algorithms, and performance characteristics. The central claim is that the package enables faithful sampling and integration across this range of mass functions.

Significance. The package fills a real need: it provides a unified, SciPy-integrated interface to many commonly used mass functions and offers both random and optimal sampling, as well as access to PMF/CMF models that are rarely available in other tools. The open-source availability and Zenodo archiving are commendable, and the paper clearly documents the implemented functional forms and the design choices (interpolation, sampling modes). However, as submitted, the implementation of at least one mass function is demonstrably incorrect, so the central claim of faithful implementations is not presently established. If the mathematical error in Eq. (6) is corrected and the interpolation accuracy is validated with concrete tests, the package would be a valuable community resource.

major comments (2)
  1. [§2.5, Eq. (6)] Equation (6) is not a valid CDF for the error-convolved power law. Differentiating the RHS does not reproduce the PDF in Eq. (5); the correct CDF is F(m)=α/(m_min^{-α}-m_max^{-α}) ∫_{m_min}^{m_max} x^{-(α+1)} Φ((m-x)/σ) dx. In the σ→0 limit, Eq. (6) approaches α(m_min^{-α}-m^{-α})/(m_min^{-α}-m_max^{-α}) for m in (m_min,m_max), i.e. α times the true CDF; for α>1 this exceeds 1 near m_max. For parameters typical of the package's examples (α=2.35, m_min=0.03, m_max=120, σ=0.5), the RHS at m=m_max is approximately 2.85, clearly inconsistent with a cumulative probability. Because KoenConvolvedPowerLaw interpolates this CDF to construct the PPF (§2.5 and §3.1), random samples from that class are drawn from the wrong distribution. The formula (and the code implementing it) must be corrected before the package can be described as faithfully implementing the KK09 distribution.
  2. [§2, interpolation paragraph] The paper states that PCHIP interpolation with the default 200-point grid keeps errors 'within ∼.1% of true value for sufficiently well-sampled functions,' but no validation is presented. This is a load-bearing claim because many central outputs (KK09, PadoanTF, Schechter, PMFs, CMFs) rely on interpolation for evaluation and sampling. Please provide measured error curves for representative functions against high-accuracy direct evaluations (e.g. using scipy.integrate.quad) as a function of the number of interpolation points, and state the conditions under which the 0.1% bound holds. Without such tests, the accuracy of all interpolated functions—and hence of sampled masses—is unquantified.
minor comments (4)
  1. [§3.2] The text 'each satisfy ∫_{b_{i-1}}^{b_i} ξ(m)dm = 1' is inconsistent with the use of the scaling factor k in Eqs. (10)–(12); it should read 'k∫_{b_{i-1}}^{b_i} ξ(m)dm = 1' (or the mass function should be explicitly rescaled). This is a clarity issue but could confuse users implementing the algorithm.
  2. [§4, Table 2] Table 2 reports single timing measurements without repeats or error bars, and the statement that 'time for both methods and memory all scale linearly with cluster mass' is not supported by any figure or fit. Please include at least three repeats with scatter, and a scaling plot or fitted exponents.
  3. [§2.1] The abstract and text use 'imf' while the PyPI project is 'initial_mass_function'; please clarify the import name and PyPI name once for users unfamiliar with the package.
  4. [§5.1.2] The description of Padoan & Nordlund (2011) CMF generation is very brief; a small pseudocode block or a sentence clarifying how masses, densities, and ages are combined to build the CMF would improve reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the package's mass-function forms are taken from independent cited literature, and its claims are about implementation, not derived from or fitted to its own outputs.

full rationale

The paper is a software/toolkit paper whose central claim is that `imf` provides working implementations of known mass functions (Salpeter, Kroupa, Chabrier, Schechter, Koen & Kondlo, Padoan & Nordlund, Hennebelle & Chabrier, McKee/Offner PMFs) as SciPy-compatible distributions. Each functional form is imported from an external, independent reference and is not derived from or fitted to the package's own outputs. There are no fitted parameters being renamed as predictions, no self-referential uniqueness theorems, and no ansatz smuggled in via self-citation. The authors' citations of their own prior work (e.g., Richardson et al. 2025, the Zenodo archive of this package itself, Ginsburg et al. 2019) are not load-bearing for the mathematical content. The paper's use of interpolation and its assertion that numerical error is ~0.1% is an unverified accuracy claim, and the reviewer's noted issue with Eq. (6) is a potential mathematical correctness problem, but neither is a circular step: the package is not 'predicting' anything from data, and its outputs are not equivalent to its inputs by construction. Therefore no circularity is present; any concerns are correctness or validation risks, not circular reasoning.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

No parameters are fitted to data in this paper. The package's defaults are drawn from the cited literature, except for the two hand-chosen numerical defaults listed above. The scientific content is inherited from earlier work, so the package contributes infrastructure rather than new physical constraints.

free parameters (2)
  • Interpolation grid size = 200 points
    Default resolution for PCHIP interpolation of expensive functions. The paper asserts <0.1% error for 'sufficiently well-sampled functions' but provides no error analysis; this choice directly controls fidelity of PMF/CMF/KK09/Schechter sampling.
  • Salpeter default mass bounds = (0.3, 120) M_sun
    The paper sets these to be 'consistent with other models,' differing from Salpeter's original (1, 100) M_sun. This is a hand-chosen default that changes the shape of the implemented Salpeter class, though it is user-tunable.
assumptions (5)
  • domain assumption The cited mass-function models (Salpeter 1955, Kroupa 2001, Chabrier 2003, Padoan & Nordlund 2002, PN2011, HC2008-13, M10/O11) are physically meaningful descriptions of the populations they target.
    The package is a tool for using these models; it does not validate them. If any are wrong, the utility of the corresponding class is limited. Invoked throughout Sections 2 and 5.
  • domain assumption Inverse-CDF sampling via SciPy rvs produces correct draws from each Distribution, including interpolated CDFs.
    Section 3.1 relies on the PPF of each distribution; for interpolated functions this assumes the PCHIP interpolation preserves quantiles well.
  • domain assumption The optimal-sampling prescription (Kroupa et al. 2013/2026, Schulz et al. 2015) is well-posed and converges under imf's iterative root-finding.
    Section 3.2; no convergence proof is given beyond bracketing and iteration.
  • domain assumption M10/O11 accretion histories and the HC/PN CMF formalisms correctly describe protostellar and core mass evolution.
    Sections 5.1.1 and 5.1.2; these are external theoretical prescriptions adopted without modification.
  • standard math scipy.integrate.quad and PCHIP interpolation compute the stated integrals and interpolants accurately for the integrands in Equations (5), (6), (13), and (16).
    The paper provides no error analysis; the claimed ~0.1% error is asserted from unpublished 'numerical testing.'

how reviews work

0 comments
Cite this review

Pith. "Pith review of The IMF package: a toolkit implementing mass functions and statistical tools to analyze them." pith.science (2026). https://pith.science/paper/RWI22HUN

@misc{pith2026260726007,
  author       = {Pith},
  title        = {Pith review of: The IMF package: a toolkit implementing mass functions and statistical tools to analyze them},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RWI22HUN}},
  note         = {Machine review of arXiv:2607.26007}
}
abstract

Mass functions are used in all areas of astrophysics. The stellar initial mass function (IMF), in particular, plays a central role in modeling stellar populations in galaxies. However, few dedicated tools for working directly with the IMF and its precursor functions are widely available. We present the $\texttt{imf}$ package, a Python library integrated into the wider scientific Python ecosystem that implements common and variant forms of mass functions, especially the IMF and its pre- and protostellar equivalents, as probability distribution functions based on SciPy's statistics architecture. This package enables the performance of operations such as sampling and integration on a wide array of highly customizable mass functions. $\texttt{imf}$ is publicly available on the Python package index $\texttt{pypi}$ under the project name $\texttt{initial_mass_function}$.

Figures

Figures reproduced from arXiv: 2607.26007 by the authors.

Figure 1
Figure 1. A sample of mass functions implemented in imf. All plotted functions span a mass range of 0.03-120 M⊙. Each IMF uses its default parameters where possible. For the error-convolved power law (§2.5), which has no defaults, parameters are included in the legend. implementation is similar to the canonical function; however, m⋆, min = 0.3M⊙ and m⋆, max = 120M⊙ by default for greater consistency with the other models. 2.2… view at source ↗
Figure 2
Figure 2. The most massive star systems in clusters of 102 /103 /104 M⊙ (left/middle/right) sampled from a Kroupa IMF according to each of imf’s methods. Random sampling (§3.1) is shown as histograms that plot the distribution of the most massive system taken from 104 cluster realizations at each mass. Optimal sampling (§3.2) is shown as a vertical line marking the most massive system possible given the cluster mass and defau… view at source ↗
Figure 3
Figure 3. Mass functions derived using the approach of P. Padoan & ˚A. Nordlund (2011). Left: CMFs sampled at the crossing time of the parent cloud tcross tracking visible cores. Transient and prestellar cores, those which have not yet collapsed, are plotted in blue while stellar cores, which have collapsed, are plotted in orange. The combined CMF is plotted in black; divided CMFs are scaled to match the total CMF. Right: The… view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: CMFs from P. Hennebelle & G. Chabrier (2008, 2009, 2013) plotted against the G. Chabrier (2003a) IMF shifted by a factor of 3 as in [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]
Figure 5
Figure 5. Figure 5 [PITH_FULL_IMAGE:figures/full_fig_p014_5.png]
Figure 6
Figure 6. Figure 6: The same as [PITH_FULL_IMAGE:figures/full_fig_p015_6.png]
Figure 7
Figure 7. Figure 7: The same as [PITH_FULL_IMAGE:figures/full_fig_p015_7.png]
Figure 8
Figure 8. Figure 8: Time evolution of total (intrinsic + accretion) luminosity for protostars of low to intermediate mass following a tapered turbulent core accretion history as described in M10. Values are derived using a version of the protostellar evolution code of M. Klassen et al. (2…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

58 extracted references · 7 canonical work pages

  1. [1]

    Alves, J., Lombardi, M., & Lada, C. J. 2007, A&A, 462, L17, doi: 10.1051/0004-6361:20066389 Astropy Collaboration, Robitaille, T. P., Tollerud, E. J., et al. 2013, A&A, 558, A33, doi: 10.1051/0004-6361/201322068 Astropy Collaboration, Price-Whelan, A. M., Sip˝ ocz, B. M., et al. 2018, AJ, 156, 123, doi: 10.3847/1538-3881/aabc4f Astropy Collaboration, Pric...

    doi:10.1051/0004-6361:20066389 2007

  2. [2]

    A., Bate, M

    Bonnell, I. A., Bate, M. R., Clarke, C. J., & Pringle, J. E. 1997, MNRAS, 285, 201, doi: 10.1093/mnras/285.1.201

    doi:10.1093/mnras/285.1.201 1997

  3. [4]

    2003a, PASP, 115, 763, doi: 10.1086/376392

    Chabrier, G. 2003a, PASP, 115, 763, doi: 10.1086/376392

    doi:10.1086/376392

  4. [5]

    2003b, ApJL, 586, L133, doi: 10.1086/374879

    Chabrier, G. 2003b, ApJL, 586, L133, doi: 10.1086/374879

    doi:10.1086/374879

  5. [6]

    2005, in Astrophysics and Space Science

    Chabrier, G. 2005, in Astrophysics and Space Science

    2005

  6. [7]

    327, The Initial Mass Function 50 Years Later, ed

    Library, Vol. 327, The Initial Mass Function 50 Years Later, ed. E. Corbelli, F. Palla, & H. Zinnecker, 41, doi: 10.1007/978-1-4020-3407-7 5

    doi:10.1007/978-1-4020-3407-7

  7. [8]

    2013, ARA&A, 51, 393, doi: 10.1146/annurev-astro-082812-141017

    Conroy, C. 2013, ARA&A, 51, 393, doi: 10.1146/annurev-astro-082812-141017

    doi:10.1146/annurev-astro-082812-141017 2013

  8. [9]

    Conroy, C., & van Dokkum, P. G. 2012, ApJ, 760, 71, doi: 10.1088/0004-637X/760/1/71 Ekstr¨ om, S., Georgy, C., Eggenberger, P., et al. 2012, A&A, 537, A146, doi: 10.1051/0004-6361/201117751

    doi:10.1088/0004-637x/760/1/71 2012

Show all 58 references
  1. [10]

    N., & Butland, J

    Fritsch, F. N., & Butland, J. 1984, SIAM Journal on Scientific and Statistical Computing, 5, 300, doi: 10.1137/0905021 17

    1984 doi

  2. [11]

    M., Brasseur, C

    Ginsburg, A., Sip˝ ocz, B. M., Brasseur, C. E., et al. 2019, AJ, 157, 98, doi: 10.3847/1538-3881/aafc33

    2019 doi

  3. [12]

    2026a, Research in Astronomy and Astrophysics, 26, 055012, doi: 10.1088/1674-4527/ae4600

    Gjergo, E., Zhang, Z., & Kroupa, P. 2026a, Research in Astronomy and Astrophysics, 26, 055012, doi: 10.1088/1674-4527/ae4600

    doi

  4. [13]

    2026b, Research in Astronomy and Astrophysics, 26, 025003, doi: 10.1088/1674-4527/ae1f79

    Gjergo, E., Zhang, Z.-Y., Kroupa, P., et al. 2026b, Research in Astronomy and Astrophysics, 26, 025003, doi: 10.1088/1674-4527/ae1f79

    doi

  5. [14]

    R., Millman, K

    Harris, C. R., Millman, K. J., van der Walt, S. J., et al. 2020, Nature, 585, 357, doi: 10.1038/s41586-020-2649-2

    2020 doi

  6. [15]

    2016, ARA&A, 54, 135, doi: 10.1146/annurev-astro-081915-023347

    Hartmann, L., Herczeg, G., & Calvet, N. 2016, ARA&A, 54, 135, doi: 10.1146/annurev-astro-081915-023347

    2016 doi

  7. [16]

    2008, ApJ, 684, 395, doi: 10.1086/589916

    Hennebelle, P., & Chabrier, G. 2008, ApJ, 684, 395, doi: 10.1086/589916

    2008 doi

  8. [17]

    2009, ApJ, 702, 1428, doi: 10.1088/0004-637X/702/2/1428

    Hennebelle, P., & Chabrier, G. 2009, ApJ, 702, 1428, doi: 10.1088/0004-637X/702/2/1428

    2009 doi

  9. [18]

    2013, ApJ, 770, 150, doi: 10.1088/0004-637X/770/2/150

    Hennebelle, P., & Chabrier, G. 2013, ApJ, 770, 150, doi: 10.1088/0004-637X/770/2/150

    2013 doi

  10. [19]

    Hennebelle, P., & Grudi´ c, M. Y. 2024, ARA&A, 62, 63, doi: 10.1146/annurev-astro-052622-031748

    2024 doi

  11. [20]

    Hopkins, P. F. 2013, MNRAS, 430, 1880, doi: 10.1093/mnras/stt010

    2013 doi

  12. [21]

    Hunter, J. D. 2007, Computing in Science & Engineering, 9, 90, doi: 10.1109/MCSE.2007.55

    2007 doi

  13. [23]

    C., Seth, A

    Johnson, L. C., Seth, A. C., Dalcanton, J. J., et al. 2017, ApJ, 839, 78, doi: 10.3847/1538-4357/aa6a1f

    2017 doi

  14. [24]

    D., Marocco, F., Gelino, C

    Kirkpatrick, J. D., Marocco, F., Gelino, C. R., et al. 2024, ApJS, 271, 55, doi: 10.3847/1538-4365/ad24e2

    2024 doi

  15. [25]

    E., & Peters, T

    Klassen, M., Pudritz, R. E., & Peters, T. 2012, MNRAS, 421, 2861, doi: 10.1111/j.1365-2966.2012.20523.x

    2012

  16. [26]

    2009, MNRAS, 397, 495, doi: 10.1111/j.1365-2966.2009.14956.x K¨ onyves, V., Andr´ e, P., Men’shchikov, A., et al

    Koen, C., & Kondlo, L. 2009, MNRAS, 397, 495, doi: 10.1111/j.1365-2966.2009.14956.x K¨ onyves, V., Andr´ e, P., Men’shchikov, A., et al. 2015, A&A, 584, A91, doi: 10.1051/0004-6361/201525861

    2009

  17. [27]

    2001, MNRAS, 322, 231, doi: 10.1046/j.1365-8711.2001.04022.x

    Kroupa, P. 2001, MNRAS, 322, 231, doi: 10.1046/j.1365-8711.2001.04022.x

    2001

  18. [28]

    2026, in Encyclopedia of Astrophysics, Volume 2, Vol

    Kroupa, P., Gjergo, E., Jerabkova, T., & Yan, Z. 2026, in Encyclopedia of Astrophysics, Volume 2, Vol. 2, 173–210, doi: 10.1016/B978-0-443-21439-4.00035-3

    2026 doi

  19. [29]

    2003, ApJ, 598, 1076, doi: 10.1086/379105

    Kroupa, P., & Weidner, C. 2003, ApJ, 598, 1076, doi: 10.1086/379105

    2003 doi

  20. [30]

    2013, in Planets, Stars and Stellar Systems

    Kroupa, P., Weidner, C., Pflamm-Altenburg, J., et al. 2013, in Planets, Stars and Stellar Systems. Volume 5: Galactic Structure and Stellar Populations, ed. T. D. Oswalt & G. Gilmore, Vol. 5 (Springer, Dordrecht), 115, doi: 10.1007/978-94-007-5612-0 4

    2013 doi

  21. [31]

    R., Fumagalli, M., da Silva, R

    Krumholz, M. R., Fumagalli, M., da Silva, R. L., Rendahl, T., & Parra, J. 2015, MNRAS, 452, 1447, doi: 10.1093/mnras/stv1374

    2015 doi

  22. [32]

    J., & Lada, E

    Lada, C. J., & Lada, E. A. 2003, ARA&A, 41, 57, doi: 10.1146/annurev.astro.41.011802.094844

    2003 arXiv

  23. [33]

    F., & Offner, S

    McKee, C. F., & Offner, S. S. R. 2010, ApJ, 716, 167, doi: 10.1088/0004-637X/716/1/167

    2010 doi

  24. [34]

    F., & Tan, J

    McKee, C. F., & Tan, J. C. 2002, Nature, 416, 59, doi: 10.1038/416059a

    2002 doi

  25. [35]

    F., & Tan, J

    McKee, C. F., & Tan, J. C. 2003, ApJ, 585, 850, doi: 10.1086/346149

    2003 doi

  26. [36]

    1998, A&A, 336, 150

    Motte, F., Andre, P., & Neri, R. 1998, A&A, 336, 150

    1998

  27. [37]

    2022, A&A, 662, A8, doi: 10.1051/0004-6361/202141677

    Motte, F., Bontemps, S., Csengeri, T., et al. 2022, A&A, 662, A8, doi: 10.1051/0004-6361/202141677

    2022 doi

  28. [38]

    Myers, P. C. 2014, ApJ, 781, 33, doi: 10.1088/0004-637X/781/1/33

    2014 doi

  29. [39]

    Offner, S. S. R., Klein, R. I., McKee, C. F., & Krumholz, M. R. 2009, ApJ, 703, 131, doi: 10.1088/0004-637X/703/1/131

    2009 doi

  30. [40]

    Offner, S. S. R., & McKee, C. F. 2011, ApJ, 736, 53, doi: 10.1088/0004-637X/736/1/53

    2011 doi

  31. [41]

    Offner, S. S. R., Moe, M., Kratter, K. M., et al. 2023, in Astronomical Society of the Pacific Conference Series, Vol. 534, Protostars and Planets VII, ed. S. Inutsuka, Y. Aikawa, T. Muto, K. Tomida, & M. Tamura, 275, doi: 10.48550/arXiv.2203.10066

    2023 doi arXiv

  32. [42]

    2002, ApJ, 576, 870, doi: 10.1086/341790

    Padoan, P., & Nordlund, ˚A. 2002, ApJ, 576, 870, doi: 10.1086/341790

    2002 doi

  33. [43]

    2011, ApJL, 741, L22, doi: 10.1088/2041-8205/741/1/L22

    Padoan, P., & Nordlund, ˚A. 2011, ApJL, 741, L22, doi: 10.1088/2041-8205/741/1/L22

    2011 doi

  34. [44]

    Padoan, P., Nordlund, A., & Jones, B. J. T. 1997, MNRAS, 288, 145, doi: 10.1093/mnras/288.1.145

    1997 doi

  35. [45]

    Palla, F., & Stahler, S. W. 1999, ApJ, 525, 772, doi: 10.1086/307928

    1999 doi

  36. [46]

    Palla, F., & Stahler, S. W. 2000, ApJ, 540, 255, doi: 10.1086/309312 Portegies Zwart, S. F., McMillan, S. L. W., & Gieles, M. 2010, ARA&A, 48, 431, doi: 10.1146/annurev-astro-081309-130834

    2000 doi

  37. [47]

    H., & Schechter, P

    Press, W. H., & Schechter, P. 1974, ApJ, 187, 425, doi: 10.1086/152650

    1974 doi

  38. [48]

    Richardson, T., Ginsburg, A., & Koposov, S. E. 2026, Zenodo, doi: 10.5281/ZENODO.21313299

    2026 doi

  39. [49]

    2025, ApJ, 989, 95, doi: 10.3847/1538-4357/ade99d

    Richardson, T., Ginsburg, A., Rosolowsky, E., Peltonen, J., & Indebetouw, R. 2025, ApJ, 989, 95, doi: 10.3847/1538-4357/ade99d

    2025 doi

  40. [50]

    Rosen, A. L. 2026, arXiv e-prints, arXiv:2603.15779, doi: 10.48550/arXiv.2603.15779

    2026 doi

  41. [51]

    Salpeter, E. E. 1955, ApJ, 121, 161, doi: 10.1086/145971 18

    1955 doi

  42. [52]

    Schneider, F. R. N., Sana, H., Evans, C. J., et al. 2018, Science, 359, 69, doi: 10.1126/science.aan0106

    2018 doi

  43. [53]

    2015, A&A, 582, A93, doi: 10.1051/0004-6361/201425296

    Schulz, C., Pflamm-Altenburg, J., & Kroupa, P. 2015, A&A, 582, A93, doi: 10.1051/0004-6361/201425296

    2015 doi

  44. [54]

    Shu, F. H. 1977, ApJ, 214, 488, doi: 10.1086/155274

    1977 doi

  45. [55]

    S., & Dav´ e, R

    Somerville, R. S., & Dav´ e, R. 2015, ARA&A, 53, 51, doi: 10.1146/annurev-astro-082812-140951

    2015 doi

  46. [56]

    D., Garmany, C

    Vacca, W. D., Garmany, C. D., & Shull, J. M. 1996, ApJ, 460, 914, doi: 10.1086/177020

    1996 doi

  47. [57]

    E., et al

    Virtanen, P., Gommers, R., Oliphant, T. E., et al. 2020, Nature Methods, 17, 261, doi: 10.1038/s41592-019-0686-2

    2020 doi

  48. [58]

    M., Johnson, L

    Wainer, T. M., Johnson, L. C., Seth, A. C., et al. 2022, ApJ, 928, 15, doi: 10.3847/1538-4357/ac51cf

    2022 doi

  49. [59]

    2006, MNRAS, 365, 1333, doi: 10.1111/j.1365-2966.2005.09824.x

    Weidner, C., & Kroupa, P. 2006, MNRAS, 365, 1333, doi: 10.1111/j.1365-2966.2005.09824.x

    2006

  50. [60]

    2017, A&A, 607, A126, doi: 10.1051/0004-6361/201730987

    Yan, Z., Jerabkova, T., & Kroupa, P. 2017, A&A, 607, A126, doi: 10.1051/0004-6361/201730987

    2017 doi

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.