Skip to content

Releases: geem-lab/overreact

v1.2.0

25 Apr 21:33
@schneiderfelipe schneiderfelipe
v1.2.0
267d097
Compare
Choose a tag to compare
v1.2.0 Latest
Latest

⚛️📈 overreact is a library and a command-line tool for building and analyzing homogeneous microkinetic models from first-principles calculations.

overreact

This is a feature release of overreact that, among other minor changes, improves integrator stability and parameter choices for the integrators overall. In particular, overreact now makes similar choices to scipy.integrate.solve_ivp, which should lead to fewer surprises. A big part of this was kindly contributed by @bmounssefjr.

⚠️ It was found out that the currently calculated analytical Jacobian leads to problems, which is most relevant when implicit integrators are employed. This affects LSODA, Radau and BDF. As such, the usage of analytical Jacobians has been disabled for this version until we find a fix for the bug. Even though it is a breaking chance per se, it is mild, as most of the change happened in places where user had previously no access. We hope to fix this issue soon. Previous and published results were not affected.

Update it using pip:

$ pip install -U 'overreact[cli,fast]'

Overreact is licensed under the MIT license.

If you use overreact in your research, please cite:

Schneider, F. S. S.; Caramori, G. F. Overreact, an in Silico Lab: Automative Quantum Chemical Microkinetic Simulations for Complex Chemical Reactions. Journal of Computational Chemistry 2022, 44 (3), 209–217. doi:10.1002/jcc.26861.

Here's the reference in BibTeX format:

@article{overreact_paper2022,
  title         = {Overreact, an in silico lab: Automative quantum chemical microkinetic simulations for complex chemical reactions},
  author        = {Schneider, Felipe S. S. and Caramori, Giovanni F.},
  year          = {2022},
  month         = {Apr},
  journal       = {Journal of Computational Chemistry},
  publisher     = {Wiley},
  volume        = {44},
  number        = {3},
  pages         = {209–217},
  doi           = {10.1002/jcc.26861},
  issn          = {1096-987x},
  url           = {http://dx.doi.org/10.1002/jcc.26861},
}

What's Changed

Full Changelog: v1.1.0...v1.2.0

Contributors

schneiderfelipe, dependabot, and bmounssefjr
Assets 2
0 Join discussion

v1.1.0

04 Jan 22:24
@schneiderfelipe schneiderfelipe
v1.1.0
74e9211
Compare
Choose a tag to compare
v1.1.0

⚛️📈 overreact is a library and a command-line tool for building and analyzing homogeneous microkinetic models from first-principles calculations.

overreact

This is a feature release of overreact that improves output, automatic choice of most parameters and adds support for Python 3.10.

Update it using pip:

$ pip install -U 'overreact[cli,fast]'

Overreact is licensed under the MIT license.

ℹ️ By the way, our manuscript has been featured in the latest issue of the Journal of Computational Chemistry! If you use overreact in your research, please cite:

Schneider, F. S. S.; Caramori, G. F. Overreact, an in Silico Lab: Automative Quantum Chemical Microkinetic Simulations for Complex Chemical Reactions. Journal of Computational Chemistry 2022, 44 (3), 209–217. doi:10.1002/jcc.26861.

Here's the reference in BibTeX format:

@article{overreact_paper2022,
  title         = {Overreact, an in silico lab: Automative quantum chemical microkinetic simulations for complex chemical reactions},
  author        = {Schneider, Felipe S. S. and Caramori, Giovanni F.},
  year          = {2022},
  month         = {Apr},
  journal       = {Journal of Computational Chemistry},
  publisher     = {Wiley},
  volume        = {44},
  number        = {3},
  pages         = {209–217},
  doi           = {10.1002/jcc.26861},
  issn          = {1096-987x},
  url           = {http://dx.doi.org/10.1002/jcc.26861},
}

What's Changed

Highlights

It just works®:
overreact is easier than ever. We fine-tuned the choice of many parameters, so that your simulations run smoothly right from the start. It is now smart enough to guess for how long to simulate (bde5bec, 3c842a7), which solver to use (928b48d), how fast should an equilibrium be (3a516bd, 1d2cbc8, 645578c), how to handle edge cases when correcting for quantum tunneling (0e5ad84), which species actually make sense to plot when generating a kinetic profile (2de3dbd), how large show the first time step be in a simulation (8eafc35, a4cd60d, ae55264), and more.

Adjust energies per species:
You can now use the extra_energy_term flag to adjust the energy of a single species (777155c).

User-interface and output:
The output has been enhanced and is clearer than ever (13fd9ae, 4f2c0d4, dec2297), which now shows the progress of the calculation as it is being performed (2d4d894). Error messages and warnings have been improved as well (92c7cb0).

Documentation and resources:
We greatly improved consistency between the documentation, API and command-line help (f0bc3b9, d828d37, 0ae4727, 1a2d030, c6daa55). In fact, we now have two sets of documentations: the user guide, focused on the use of the command-line application, and the Python API documentation (b1c7456, 4ecacfc, ea298d0, 8ca1c7f, b6c8886, da22151, 8eb5f06, 0212376). We now have an improved README.md file (79ba5f2, df90408, c2b62ed, 0be2ad8, 64da6b9, #149), as well as a great CONTRIBUTING.md thanks to @Leticia-maria (fcccbd1, f1ce7c0, #133). Thanks, Letícia!

Code quality and API:
The user-facing API is now almost completely type-hinted (37cca8d, 819773c). There might be some rough edges yet, so feel free to open an issue if you see something. Lots of housekeeping has been taken place as well, mostly to keep the code base modern, performant, and easy to use (f47206b, ea27d5b, 2f6a176, fb99f6c, 7622789, f60a705). We now target Python versions 3.8-3.10 (bc67e7d, c089244), and use Ruff and Black to keep the code base as tidy as possible (cddc09d, a0f1be3, 14dde86).

Pull requests

Here are the list of relevant pull requests since last version.

New Contributors

Full Changelog: v1.0.2...v1.1.0

Contributors

dependabot and Leticia-maria
Assets 2
0 Join discussion

v1.0.2

26 Nov 21:49
@schneiderfelipe schneiderfelipe
v1.0.2
0f1dafc
Compare
Choose a tag to compare
v1.0.2

This is a patch release of overreact that solves a bug in the simulation code that had to do with the treatment of equilibrium reactions.

Update it using pip:

$ pip installl -U "overreact[cli,fast]"

overreact is licensed under the MIT license.

If you use overreact in your research, please cite our Zenodo DOI for now. A manuscript is currently in preparation.

Changes

  • Fixed a bug that led to unadjusted reaction rate constants for some half-equilibria reactions (7db8feb)
  • Added code coverage at Codecov (#101)
  • Fixed GitHub actions, which were always failing for pull requests (#106)
  • Minor changes in the CLI output header (15dc14d)
Assets 2
0 Join discussion

v1.0.1

03 Nov 19:51
@schneiderfelipe schneiderfelipe
v1.0.1
2456787
Compare
Choose a tag to compare
v1.0.1

That is the second release of overreact. A manuscript about it is in preparation, but a Zenodo DOI is already available for citations 🎉.

Install it using pip:

$ pip installl "overreact[cli,fast]"

overreact is licensed under the MIT license.

Notable changes

This is a small patch to fix and change some aspects of the project. See changes below.

Assets 2
0 Join discussion

v1.0alpha

02 Aug 20:52
@schneiderfelipe schneiderfelipe
v1.0alpha
d51e1b0
This commit was created on GitHub.com and signed with GitHub’s verified signature. The key has expired.
GPG key ID: 4AEE18F83AFDEB23
Expired
Learn about vigilant mode.
Compare
Choose a tag to compare
v1.0alpha Pre-release
Pre-release

That is the first release of overreact, still alpha, but very functional. A manuscript about it is in preparation, but a Zenodo DOI will soon be available for citations.

overreact is licensed under the MIT license (#12).

You can read the documentation at https://geem-lab.github.io/overreact-docs/.

Some notable features

  • Well-tested thermodynamics (#5, #7, #17, #25, #64, abb2417)
    • Automatic support of molecular symmetries (#8, #62)
    • Automatic inversion of small imaginary frequencies, with a warning (#50)
    • Direct bias adjustment of computed free energies for fitting concentration profiles (#13)
  • Thoroughly tested (#45, #53, #54, #57)
  • Good command-line interface (#41, #47, #66, #67, #73)
  • Stable numerical integration of the reaction ODE system (#48, #55, #68, #69)
    • Support for fixing concentration in chemical kinetic models (#14, #74)
  • Good documentation (#49, #83, #87)
  • Semantic versioning (#43)
Assets 2
0 Join discussion