Pith. sign in

REVIEW 2 major objections 3 minor 1 cited by

SimplySQS claims that special quasirandom structure (SQS) generation with ATAT's mcsqs can be automated through an online interface that emits a single shareable all-in-one bash script, and that the workflow, applied to Pb1-xSrxTiO3, reprod

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-04 08:53 UTC pith:PVTKU3NP

load-bearing objection Useful and mostly sound SQS workflow wrapper with a genuine contradiction in the case study and an overstrong reproducibility claim. the 2 major comments →

arxiv 2510.18020 v4 pith:PVTKU3NP submitted 2025-10-20 cond-mat.mtrl-sci

SimplySQS: An Automated and Reproducible Workflow for Special Quasirandom Structure Generation with ATAT

Miroslav Lebeda , Jan Drahokoupil , Petr Vl\v{c}\'ak , \v{S}imon Svoboda , Axel van de Walle This is my paper
classification cond-mat.mtrl-sci
keywords special quasirandom structuresATAT mcsqsdisordered alloysreproducible workflowWarren-Cowley parametersrandomness scoreperovskitemachine-learning interatomic potential
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper introduces SimplySQS, an online interface that standardizes the otherwise manual workflow of generating special quasirandom structures (SQSs)—periodic supercells designed to mimic the local atomic correlations of a random alloy—using the ATAT mcsqs engine. The central claim is that the entire process, from structure import and composition definition to cluster-parameter selection, input-file creation, search execution, and output conversion, can be bundled into a single shareable bash script, making SQS generation reproducible rather than user-dependent. To show this matters, the authors generate SQSs for the full Pb1-xSrxTiO3 composition range and relax them with a universal machine-learning interatomic potential; the computed lattice parameters reproduce the known cubic-to-tetragonal phase transition near x ≈ 0.5, with deviations under about 1% in the cubic region and under 4% in the tetragonal region. The paper also introduces a normalized randomness score, based on Warren-Cowley short-range order parameters, that tells users whether an SQS search is needed or random occupations already suffice. If correct, SimplySQS lowers the barrier to using a mature, widely used SQS tool and makes computed disordered-alloy structures easier to share, reproduce, and teach.

Core claim

The discovery is procedural: a browser-based workflow can turn ATAT's mcsqs, a console-based special quasirandom structure generator, into a guided pipeline whose output is a single 'monitor.sh' script. That script recreates the required input files, launches the mcsqs search (single or parallel, or as a concentration sweep), prints and logs the objective-function progress, and on termination converts the best SQS from ATAT's internal 'bestsqs.out' format into standard POSCAR, from which VASP, LAMMPS, and other codes can proceed directly. The paper further claims that the same script can be shared and rerun on different computers, making the search reproducible. The accompanying PSTO case st

What carries the argument

The load-bearing object is the auto-generated all-in-one bash script ('monitor.sh'), which encapsulates the entire mcsqs search: it writes 'rndstr.in' and 'sqscell.out', runs mcsqs with user-chosen cluster cutoffs, number of parallel runs, and time budget, streams real-time objective-function progress to the console, and converts the final 'bestsqs.out' into POSCAR. A second mechanism is the normalized randomness score: for a given supercell and composition, the workflow generates many random configurations, computes Warren-Cowley short-range order parameters for the first three coordination shells on each sublattice, combines their deviations from ideal randomness into a score from 0 to 1,

Load-bearing premise

The load-bearing premise is that rerunning the same generated script on a different computer reproduces the exact same final structure; the paper does not show that a random seed is fixed or that software versions are pinned, so this premise, if false, would undo the reproducibility claim.

What would settle it

Run the identical 'monitor.sh' script on two machines with different ATAT installations or library versions and compare the objective-function value and atomic coordinates in 'bestsqs.out'; if they differ, the claim that the search is reliably reproducible on different computers is false.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • SQS generation becomes a one-file operation: a user only needs to share 'monitor.sh' plus the input structure for another researcher to recreate the same search.
  • The concentration-sweep mode makes full composition-range alloy models practical, since one script can loop over many x values, run parallel searches for each, and convert each best result to a simulation-ready format.
  • The randomness score gives a cheap diagnostic for whether random atomic occupation is adequate or SQS optimization is needed, potentially saving substantial compute on simple systems.
  • The PSTO demonstration suggests that coupling automated SQS generation with universal machine-learning potentials is a viable pipeline for mapping composition-dependent lattice parameters and phase transitions.
  • Standardized conversion to POSCAR, CIF, LMP, and XYZ removes the most common source of user error in ATAT workflows and connects directly to DFT and molecular dynamics packages.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The paper stops short of proving reproducibility across machines; a natural extension would be a containerized or pinned-version execution environment so the script's deterministic claim is actually testable.
  • If the script is deterministic, SQS searches could be archived as reproducible artifacts alongside publications, letting reviewers re-run the exact search rather than trusting the reported objective function.
  • The randomness-score thresholds (0.95 and 0.85) are heuristic; they could be calibrated by comparing predicted 'sufficiently random' supercells against downstream property convergence in DFT.
  • The same workflow pattern—web form, generated script, output parsing—could be adapted to other SQS engines beyond ATAT, making reproducibility comparisons between tools easier.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 3 minor

Summary. The paper presents SimplySQS, a web-based Streamlit application that automates the preparation, execution, and post-processing of special quasirandom structure (SQS) calculations with the ATAT mcsqs module. The tool guides users through structure import, composition/supercell definition, cluster-parameter selection, and generates ATAT input files plus an all-in-one bash script (monitor.sh) intended to make SQS searches reproducible and easy to run. The workflow also includes a randomness-quality assessment based on Warren–Cowley parameters and a binary concentration sweep mode. As a demonstration, the authors generate SQSs for Pb1-xSrxTiO3 across the full composition range with 320-atom supercells, relax them with the MACE MATPES-r2SCAN-0 machine-learning interatomic potential, and compare the resulting lattice parameters with experiment, claiming to reproduce a cubic-to-tetragonal transition near x ≈ 0.5.

Significance. If the claims are substantiated, SimplySQS would be a genuinely useful community tool: it lowers the barrier to using ATAT, standardizes a workflow that is often ad hoc, and provides logging and conversion features that many users would benefit from. The PSTO case study is a nontrivial test of the workflow and uses a publicly available code and example files, which is a strength. However, the submitted manuscript has two load-bearing problems: an internal contradiction about which composition range is cubic versus tetragonal in the central case study, and an unsupported assertion that the all-in-one script ensures exact reproducibility across computers despite no evidence of fixed/recorded random seeds or pinned software versions. These issues must be resolved before the paper's central claims can be accepted.

major comments (2)
  1. [§4.3, Abstract, §6] There is a direct internal contradiction in the central result. The abstract states that the cubic region is x > 0.5 and the tetragonal region is x < 0.5, and §6 repeats this. But §4.3 states the opposite: 'For x < 0.5, all structures remained cubic (a = c)' and 'For x > 0.5, a tetragonal distortion emerged' with a decreasing and c increasing. Since x is the Sr content, only one mapping can be consistent with the experimentally known phase behavior and with the paper's own claim of reproducing a transition near x ≈ 0.5. The reported deviations also change meaning between the two mappings (0.3% vs <1% in the cubic region; 3.4% vs <4% in the tetragonal region). This must be corrected consistently in the abstract, §4.3, §6, and Figure 10, or the demonstration of the phase transition is not interpretable.
  2. [§3.2, §6] The paper's headline reproducibility claim is not supported by evidence. §3.2 states that the all-in-one script 'can be easily shared and rerun on different computers, ensuring that the SQS search process can be reliably reproduced,' and §6 repeats that the script 'ensures reproducibility.' Yet the manuscript does not show that the script fixes or records random seeds, does not state that ATAT/Python package versions are pinned, and even notes that parallel mode launches independent searches 'with different initial seeds,' implying that seed choice is varied at runtime. Since mcsqs is stochastic, rerunning under these conditions will generally produce different SQS configurations unless a master seed is set or logged. The authors should either provide concrete evidence of seed handling and version pinning (e.g., show the relevant part of monitor.sh, add a reproducibility statement), or w
minor comments (3)
  1. [§3.4] The randomness-score thresholds (0.95, 0.85) are presented without justification. A brief explanation of how these cutoff values were chosen, or a reference to a standard, would help readers interpret the classification.
  2. [References and citations] Several citation numbers appear mismatched. In §4.2, the r²SCAN functional is cited as [33] and PBE as [34], but in the reference list [33] is Longo et al. and [34] is Xing et al. Similarly, §3.4 cites Warren–Cowley parameters as [31,32], which in the list are Furness et al. and Perdew et al. Please recheck all numeric citations against the bibliography.
  3. [§4.1] The phrase '5 parallely running concentrations' should read '5 concentrations running in parallel' or similar. Also, the cluster cutoff is described as '1.5 Å' and '1.2 Å' but immediately qualified as 'in normalized unit cell units'; the units should be stated consistently and clearly, since mcsqs uses normalized lattice parameters, not Å.

Circularity Check

0 steps flagged

No circular derivation: the PSTO result is an external MLIP benchmark against experiment, and ATAT/MACE citations are independent support for the wrapped workflow.

full rationale

No load-bearing circular step is present. SimplySQS is a workflow wrapper around ATAT's mcsqs; the paper does not claim to derive SQS theory or mcsqs behavior from its own tool. The PSTO demonstration is an independent application: SQS structures generated by the wrapped ATAT code are relaxed with MACE MATPES-r2SCAN-0, an externally trained universal machine-learning interatomic potential, and the resulting lattice parameters are compared with experimental reference data. The predictions therefore do not reduce to the paper's own inputs by construction. The ATAT citations include a co-author, but this is appropriate and non-circular because mcsqs is the wrapped third-party software and the cited implementation is externally reproduced and widely used. The randomness score is definitional (one minus averaged RMS Warren-Cowley deviation), but it is presented as a diagnostic metric with heuristic thresholds, not as a predicted physical quantity. The reproducibility claim about the shared all-in-one script is an empirical assertion about determinism and seeding that is not supported in the paper; this is a correctness risk, not circularity. Overall, the central claims are self-contained against external benchmarks and no fitted parameter is renamed as a prediction.

Axiom & Free-Parameter Ledger

3 free parameters · 5 axioms · 0 invented entities

The paper does not introduce new physical entities. Its free parameters are mostly user-chosen settings for the demo and the hand-set thresholds for the randomness score. The main assumptions are the correctness of the wrapped tools (ATAT, MACE) and the adequacy of the SQS supercell settings.

free parameters (3)
  • Randomness score classification thresholds = 0.95 and 0.85
    Chosen by hand to classify structures as sufficiently random, marginal, or ordered. No derivation or sensitivity analysis is provided in §3.4.
  • Cluster cutoff distances for PSTO demo = pair 1.5 Å, triplet 1.2 Å (normalized cell units)
    User-specified in the binary concentration sweep for Pb1-xSrxTiO3; the paper does not justify that these cutoffs capture all relevant correlations for the random alloy.
  • Sequential initialization of lattice parameters = previous lower-Pb composition
    In §4.2, each composition starts from the optimized lattice of the prior step, which could bias the optimization toward the previous phase (hysteresis). This choice is a modeling assumption.
axioms (5)
  • domain assumption ATAT mcsqs correctly generates special quasirandom structures according to the cluster expansion formalism.
    The entire tool relies on the correctness of mcsqs, which is cited as [11,12] but not re-verified here.
  • domain assumption The MACE MATPES-r2SCAN-0 universal machine-learning potential provides sufficiently accurate lattice parameters for Pb1-xSrxTiO3 across the full composition range.
    The paper's quantitative demo depends on this. The authors themselves note deviations up to 3.4% and attribute them to training data, so the accuracy is not universal.
  • domain assumption A 320-atom 4×4×4 supercell with the chosen cluster cutoffs is a faithful representation of the random alloy.
    The SQS quality is only indirectly checked via the objective function plateau; no verification that the final SQS correlations match the ideal random alloy within a tolerance is shown.
  • domain assumption The experimental reference data [36] for PSTO lattice parameters is accurate and representative.
    The comparison is made against this single reference, but the paper does not discuss experimental uncertainty or sample preparation.
  • domain assumption Fixing cell angles and enforcing a=b during relaxation does not qualitatively affect the phase transition behavior.
    This constraint in §4.2 could suppress small monoclinic or orthorhombic distortions, but the paper only compares a and c.

pith-pipeline@v1.3.0-alltime-deepseek · 10520 in / 12342 out tokens · 99628 ms · 2026-08-04T08:53:46.782027+00:00 · methodology

0 comments
read the original abstract

The special quasirandom structure (SQS) method is widely used for modeling disordered materials under periodic boundary conditions, with the ATAT mcsqs module being one of the most established implementations. However, SQS generation with mcsqs typically relies on manual preparation of input files, ad hoc execution scripts, and post-processing steps, which introduces user-dependent errors and limits reproducibility. Here, we present SimplySQS (https://simplysqs.com), an automated and reproducible workflow for SQS generation that is delivered through an online, interactive interface. SimplySQS guides users through structure import, compositional and supercell definition, and cluster parameter selection, while automatically generating all required ATAT input files and a single all-in-one execution script that encapsulates the complete search process. By standardizing input preparation, execution, and output analysis, the framework minimizes errors associated with manual file handling and enables consistent reproducibility of SQS searches. The workflow is demonstrated on the Pb1-xSrxTiO3 (PSTO, including PbTiO3 (PTO) and SrTiO3 (STO)) perovskite system. SQSs spanning the entire concentration range were generated using a single automated bash script produced by SimplySQS, after which all resulting structures were subjected to geometry optimization using a universal machine-learning interatomic potential (MACE MATPES-r2SCAN-0). This approach reliably reproduced the experimentally observed cubic-to-tetragonal transition near x = 0.5, with lattice parameters deviating by less than 1 % in the cubic region (x > 0.5) and less than 4 % in the tetragonal region (x < 0.5). Overall, SimplySQS transforms SQS generation with ATAT into an intuitive, reproducible, and systematic framework for modeling disordered materials.

Figures

Figures reproduced from arXiv: 2510.18020 by Axel van de Walle, Jan Drahokoupil, Miroslav Lebeda, Petr Vl\v{c}\'ak, \v{S}imon Svoboda.

Figure 6
Figure 6. Figure 6: Interface of the binary concentration sweep mode showing the configurable parameters: selected mcsqs mode, total search time, number of parallel runs per concentration, number of concentrations executed simultaneously, and the update interval for printing progress information to the console. 3.3 Analysis of ATAT mcsqs Outputs The SimplySQS interface also facilitates post-processing and evaluation of result… view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

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

  1. Revealing interstitial energetics in Ti-23Nb-0.7Ta-2Zr gum metal base alloy via universal machine learning interatomic potentials

    cond-mat.mtrl-sci 2025-12 conditional novelty 6.0

    A 6,750-site screening with three universal machine-learning potentials maps C/N/O/H interstitial energetics in Ti-23Nb-0.7Ta-2Zr, finding Ti stabilizes and Nb destabilizes interstitials, with SevenNet-0 failing on H ...

Reference graph

Works this paper leans on

46 extracted references · 1 linked inside Pith · cited by 1 Pith paper

  1. [1]

    One simple and transparent approach is to construct a large supercell and randomly assign elements to atomic sites

    Introduction Accurate and efficient modeling of disordered alloys is essential in materials science, as many technologically significant materials adopt disordered structures. One simple and transparent approach is to construct a large supercell and randomly assign elements to atomic sites. However, due to the finite size of the supercell, this method oft...

  2. [2]

    It is currently hosted on the free Streamlit community cloud server, which provides direct deployment from the public GitHub repository ( github.com/bracerino/atat-sqs-gui)

    Implementation Details The SimplySQS is implemented in Python and deployed online (simplysqs.com) as an interactive web interface using the Streamlit framework (streamlit.io). It is currently hosted on the free Streamlit community cloud server, which provides direct deployment from the public GitHub repository ( github.com/bracerino/atat-sqs-gui). The rep...

  3. [3]

    Application Workflow and Features SimplySQS is designed to guide users through the entire workflow of preparing input files, running the SQS search via the ATAT mcsqs with a comprehensive all-in-one bash script, and analyzing output files (see the general workflow in Fig. 1). Currently, the main features of the application include: • Structure import : Up...

  4. [4]

    Automatically creating input files with the specified parameters

  5. [5]

    Starting the mcsqs search with the chosen cluster settings

  6. [6]

    Printing real -time progress to the console, including elapsed runtime and best objective function values as well as creating a new file which contains time - dependent progress of objective function (‘mcsqs_progress.csv’ for a single run or ‘mcsqs_parallel_progress.csv’ for a parallel run)

  7. [7]

    Allowing users to set a maximum runtime for the mcsqs search and, upon completion or manual termination, automatically converting the final ‘bestsqs.out’ SQS format into the POSCAR format, ensuring direct compatibility with VASP and other simulation packages. An additional advantage of the all -in-one script is that it directly facilitates reproducibility...

  8. [8]

    Each SQS comprised a 320 -atom supercell, constructed as a 4 × 4 × 4 expansion of the initial PbTiO3 unit cell

    Example – Pb1-xSrxTiO3 Lattice Parameters with MACE r2SCAN Universal MLIP 4.1 SQS Settings To illustrate the effectiveness of SimplySQS, a series of SQSs was generated for the perovskite alloy system Pb1-xSrxTiO3 (PSTO, including PbTiO3 (PTO) and SrTiO3 (STO)) covering the full composition range (0 < x < 1). Each SQS comprised a 320 -atom supercell, const...

  9. [9]

    Impact on User Adoption and Education The presented interactive interface for preparing ATAT mcsqs input files significantly lowers the entry barrier for generating SQS with ATAT mcsqs, particularly for users with limited programming or command -line experience. Traditionally, the use of ATAT has required manual preparation of input files, familiarity wit...

  10. [10]

    Conclusions In this work, we introduced SimplySQS (simplysqs.com), an automated and reproducible workflow for generating and analyzing special quasirandom structures using the ATAT mcsqs module. The online, interactive interface guides users through the complete process, from structure import and supercell setup to automated generation of ATAT input files...

  11. [11]

    SGS24/121/OHK2/3T/12]

    Acknowledgments This work was supported by the Grant Agency of the Czech Technical University in Prague [grant No. SGS24/121/OHK2/3T/12]

  12. [12]

    The underlying code is publicly available on the GitHub repository , containing documentation for the local installation: https://github.com/bracerino/atat-sqs-gui

    Code Availability The online version of SimplySQS is accessible at: https://simplysqs.com. The underlying code is publicly available on the GitHub repository , containing documentation for the local installation: https://github.com/bracerino/atat-sqs-gui. The all-in-one bash script generated by SimplySQS and used in the Pb1-xSrxTiO3 case study, together w...

  13. [13]

    Zunger, S.-H

    A. Zunger, S.-H. Wei, L. G. Ferreira, and J. E. Bernard, Special quasirandom structures, Phys Rev Lett 65, 353 (1990)

    1990

  14. [14]

    Vizoso and C

    D. Vizoso and C. Deo, Determination of vacancy formation energies in binary UZr alloys using special quasirandom structure methods, Front Mater 8, 692660 (2021)

    2021

  15. [15]

    J. Hu, X. Zhang, C. Hu, J. Sun, X. Wang, H.-Q. Lin, and G. Li, Correlated flat bands and quantum spin liquid state in a cluster Mott insulator, Commun Phys 6, 172 (2023)

    2023

  16. [16]

    T. Wang, W. Li, C. Ni, and A. Janotti, Band gap and band offset of Ga 2 O 3 and (Al x Ga 1− x) 2 O 3 alloys, Phys Rev Appl 10, 011003 (2018)

    2018

  17. [17]

    A. M. Ritzmann, A. B. Muñoz-García, M. Pavone, J. A. Keith, and E. A. Carter, Ab initio DFT+ U analysis of oxygen vacancy formation and migration in La1-xSr x FeO3-δ (x= 0, 0.25, 0.50), Chemistry of Materials 25, 3011 (2013)

    2013

  18. [18]

    W. Hu, M. Pan, Z. Huang, H. Yang, H. Wu, L. Pan, Z. Cao, and Y . Zhao, The Stability and Elasticity in Ta-Ti-V Medium-Entropy Alloys Using First-Principles Calculations, Front Mater 8, 651435 (2021)

    2021

  19. [19]

    K. Li, X. Hu, R. Guo, W. Jiang, L. Zeng, L. Li, P . Yu, K. Wang, C. Zhang, and S. Guo, Superconductivity in the medium-entropy alloy TiVNbTa with a body-centered cubic structure, The Journal of Physical Chemistry C 127, 16211 (2023)

    2023

  20. [20]

    J. Sun, S. Mukhopadhyay, A. Subedi, T. Siegrist, and D. J. Singh, Transport properties of cubic crystalline Ge2Sb2Te5: A potential low-temperature thermoelectric material, Appl Phys Lett 106, (2015)

    2015

  21. [21]

    J. Han, J. Park, M. Kim, S. Lee, J. M. Heo, Y . H. Jin, Y . Chae, J. Han, J. Wang, and S. Seok, Ultrahigh Conductive MXene Films for Broadband Electromagnetic Interference Shielding, Advanced Materials 2502443 (2025)

    2025

  22. [22]

    P . P . P . O. Borges, R. O. Ritchie, and M. Asta, Chemical trends favoring interstitial cluster formation in bcc high-entropy alloys from first-principles calculations, Acta Mater 121091 (2025)

    2025

  23. [23]

    van de Walle, P

    A. van de Walle, P . Tiwary, M. de Jong, D. L. Olmsted, M. Asta, A. Dick, D. Shin, Y . Wang, L.-Q. Chen, and Z.-K. Liu, Efficient stochastic generation of special quasirandom structures, Calphad 42, 13 (2013)

    2013

  24. [24]

    Van De Walle, M

    A. Van De Walle, M. Asta, and G. Ceder, The alloy theoretic automated toolkit: A user guide, Calphad 26, 539 (2002)

    2002

  25. [25]

    Ångqvist, W

    M. Ångqvist, W. A. Muñoz, J. M. Rahm, E. Fransson, C. Durniak, P . Rozyczko, T. H. Rod, and P . Erhart, ICET–a Python library for constructing and sampling alloy cluster expansions, Adv Theory Simul 2, 1900015 (2019)

    2019

  26. [26]

    Gehringer, M

    D. Gehringer, M. Friák, and D. Holec, Models of configurationally-complex alloys made simple, Comput Phys Commun 286, 108664 (2023)

    2023

  27. [27]

    S. P . Ong, W. D. Richards, A. Jain, G. Hautier, M. Kocher, S. Cholia, D. Gunter, V . L. Chevrier, K. A. Persson, and G. Ceder, Python Materials Genomics (pymatgen): A robust, open-source python library for materials analysis, Comput Mater Sci 68, 314 (2013)

    2013

  28. [28]

    A. H. Larsen, J. J. Mortensen, J. Blomqvist, I. E. Castelli, R. Christensen, M. Dułak, J. Friis, M. N. Groves, B. Hammer, and C. Hargus, The atomic simulation environment— a Python library for working with atoms, Journal of Physics: Condensed Matter 29, 273002 (2017)

    2017

  29. [29]

    Batatia, P

    I. Batatia, P . Benner, Y . Chiang, A. M. Elena, D. P . Kovács, J. Riebesell, X. R. Advincula, M. Asta, M. Avaylon, and W. J. Baldwin, A foundation model for atomistic materials chemistry, ArXiv Preprint ArXiv:2401.00096 (2023)

    Pith/arXiv arXiv 2023

  30. [30]

    L. Ward, A. Dunn, A. Faghaninia, N. E. R. Zimmermann, S. Bajaj, Q. Wang, J. Montoya, J. Chen, K. Bystrom, and M. Dylla, Matminer: An open source toolkit for materials data mining, Comput Mater Sci 152, 60 (2018)

    2018

  31. [31]

    C. R. Harris, K. J. Millman, S. J. Van Der Walt, R. Gommers, P . Virtanen, D. Cournapeau, E. Wieser, J. Taylor, S. Berg, and N. J. Smith, Array programming with NumPy, Nature 585, 357 (2020)

    2020

  32. [32]

    McKinney, pandas: a foundational Python library for data analysis and statistics, Python for High Performance and Scientific Computing 14, 1 (2011)

    W. McKinney, pandas: a foundational Python library for data analysis and statistics, Python for High Performance and Scientific Computing 14, 1 (2011)

    2011

  33. [33]

    Rego and D

    N. Rego and D. Koes, 3Dmol. js: molecular visualization with WebGL, Bioinformatics 31, 1322 (2015)

    2015

  34. [34]

    J. D. Hunter, Matplotlib: A 2D graphics environment, Comput Sci Eng 9, 90 (2007)

    2007

  35. [35]

    Gražulis, D

    S. Gražulis, D. Chateigner, R. T. Downs, A. F . T. Yokochi, M. Quirós, L. Lutterotti, E. Manakova, J. Butkus, P . Moeck, and A. Le Bail, Crystallography Open Database–an open-access collection of crystal structures, Applied Crystallography 42, 726 (2009)

    2009

  36. [36]

    M. K. Horton, P . Huck, R. X. Yang, J. M. Munro, S. Dwaraknath, A. M. Ganose, R. S. Kingsbury, M. Wen, J. X. Shen, and T. S. Mathis, Accelerated data-driven materials science with the Materials Project, Nat Mater 1 (2025)

    2025

  37. [37]

    A. Jain, S. P . Ong, G. Hautier, W. Chen, W. D. Richards, S. Dacek, S. Cholia, D. Gunter, D. Skinner, and G. Ceder, Commentary: The Materials Project: A materials genome approach to accelerating materials innovation, APL Mater 1, (2013)

    2013

  38. [38]

    S. P . Huber, M. Minotakis, M. Bercx, T. Reents, K. Eimre, N. Paulish, N. Hörmann, M. Uhrin, N. Marzari, and G. Pizzi, MC3D: The materials cloud computational database of experimentally known stoichiometric inorganics, ArXiv Preprint ArXiv:2508.19223 (2025)

    arXiv 2025

  39. [39]

    C. W. Andersen, R. Armiento, E. Blokhin, G. J. Conduit, S. Dwaraknath, M. L. Evans, Á. Fekete, A. Gopakumar, S. Gražulis, and A. Merkys, OPTIMADE, an API for exchanging materials data, Sci Data 8, 217 (2021)

    2021

  40. [40]

    Hafner, Ab‐initio simulations of materials using VASP: Density‐functional theory and beyond, J Comput Chem 29, 2044 (2008)

    J. Hafner, Ab‐initio simulations of materials using VASP: Density‐functional theory and beyond, J Comput Chem 29, 2044 (2008)

    2044

  41. [41]

    Kresse and J

    G. Kresse and J. Furthmüller, Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set, Phys Rev B 54, 11169 (1996)

    1996

  42. [42]

    A. P . Thompson, H. M. Aktulga, R. Berger, D. S. Bolintineanu, W. M. Brown, P . S. Crozier, P . J. In’t Veld, A. Kohlmeyer, S. G. Moore, and T. D. Nguyen, LAMMPS-a flexible simulation tool for particle-based materials modeling at the atomic, meso, and continuum scales, Comput Phys Commun 271, 108171 (2022)

    2022

  43. [43]

    J. W. Furness, A. D. Kaplan, J. Ning, J. P . Perdew, and J. Sun, Accurate and numerically efficient r2SCAN meta-generalized gradient approximation, J Phys Chem Lett 11, 8208 (2020)

    2020

  44. [44]

    J. P . Perdew, K. Burke, and M. Ernzerhof, Generalized gradient approximation made simple, Phys Rev Lett 77, 3865 (1996)

    1996

  45. [45]

    V . M. Longo, M. das G. S. Costa, A. Z. Simoes, I. L. V . Rosa, C. O. P . Santos, J. Andres, E. Longo, and J. A. Varela, On the photoluminescence behavior of samarium-doped strontium titanate nanostructures under UV light. A structural and electronic understanding, Physical Chemistry Chemical Physics 12, 7566 (2010)

    2010

  46. [46]

    X. Xing, J. Chen, J. Deng, and G. Liu, Solid solution Pb1− xSrxTiO3 and its thermal expansion, J Alloys Compd 360, 286 (2003)

    2003