Docs: suggesting smaller mixing parameters for non converging spin-polarized calculations#7589
Open
Cstandardlib wants to merge 2 commits into
Open
Docs: suggesting smaller mixing parameters for non converging spin-polarized calculations#7589Cstandardlib wants to merge 2 commits into
Cstandardlib wants to merge 2 commits into
Conversation
…scf hard to converge.
There was a problem hiding this comment.
Pull request overview
Updates ABACUS input-parameter documentation (and the corresponding built-in help text) to advise smaller/paired mixing parameters for hard-to-converge spin-polarized SCF calculations, in response to issue #7125.
Changes:
- Extend
mixing_betahelp text to recommend reducingmixing_betaandmixing_beta_magtogether for difficult spin-polarized SCF. - Extend
mixing_beta_maghelp text with the same paired-reduction guidance. - Update
docs/parameters.yamlmixing sections accordingly (but also includes multiple unrelated documentation changes beyond the stated PR scope).
Reviewed changes
Copilot reviewed 2 out of 2 changed files in this pull request and generated 11 comments.
| File | Description |
|---|---|
| source/source_io/module_parameter/read_input_item_elec_stru.cpp | Updates built-in parameter descriptions for mixing_beta and mixing_beta_mag. |
| docs/parameters.yaml | Updates parameter docs for mixing guidance and includes several additional documentation edits across other parameters. |
Comment on lines
579
to
+583
| * 0.1 or less: if convergence of SCF calculation is difficult to reach, please try 0 < mixing_beta < 0.1. | ||
|
|
||
| Note: For low-dimensional large systems, the setup of mixing_beta=0.1, mixing_ndim=20, and mixing_gg0=1.0 usually works well.)"; | ||
| Note: For low-dimensional large systems, the setup of mixing_beta=0.1, mixing_ndim=20, and mixing_gg0=1.0 usually works well. | ||
|
|
||
| For spin-polarized calculations (nspin=2 or nspin=4) that are difficult to converge, try reducing both mixing_beta and mixing_beta_mag simultaneously, e.g., mixing_beta=0.1 and mixing_beta_mag=0.1 or lower.)"; |
Comment on lines
+616
to
+619
| item.description = R"(Mixing parameter of magnetic density. | ||
|
|
||
| If SCF convergence is difficult with spin polarization (nspin=2 or nspin=4), try reducing both mixing_beta and mixing_beta_mag simultaneously, e.g., mixing_beta=0.1 and mixing_beta_mag=0.1 or lower.)"; | ||
|
|
| type: Real | ||
| description: | | ||
| It's the density threshold for electronic iteration. It represents the charge density error between two sequential densities from electronic iterations. This criterion is always enabled. If `scf_ene_thr` is set, the total-energy criterion (`scf_ene_thr`) is evaluated conditionally after the charge-density criterion (`scf_thr`) is satisfied, and not on the first iteration. For local-orbital calculations, 1e-6 is usually accurate enough. | ||
| It's the density threshold for electronic iteration. It represents the charge density error between two sequential densities from electronic iterations. Usually for local orbitals, usually 1e-6 may be accurate enough. |
Comment on lines
1395
to
1399
| * nhc: Nose-Hoover chain, see md_tfreq and md_tchain in detail. | ||
| * anderson: Anderson thermostat, see md_nraise in detail. | ||
| * berendsen: Berendsen thermostat, see md_nraise in detail. | ||
| * rescaling: velocity Rescaling method 1, see md_tolerance in detail. | ||
| * rescale_v: velocity Rescaling method 2, see md_nraise in detail. |
| type: Boolean | ||
| description: | | ||
| Whether to print Hamiltonian matrices H(R) in npz format. The output files are named output_HR0.npz, output_HR1.npz, and so on according to spin channel. This feature requires ABACUS to be built with CNPY. | ||
| Whether to print Hamiltonian matrices H(R) in npz format. This feature does not work for gamma-only calculations. |
| availability: sc_mag_switch is true | ||
| - name: sc_scan_lambda_start | ||
| category: Spin-Constrained DFT | ||
| type: Float |
| availability: sc_lambda_strategy is linear_scan | ||
| - name: sc_scan_lambda_end | ||
| category: Spin-Constrained DFT | ||
| type: Float |
Comment on lines
+140
to
+153
| - name: assume_isolated | ||
| category: System variables | ||
| type: String | ||
| description: | | ||
| Used to perform a calculation assuming an isolated system in a 3D supercell. | ||
|
|
||
| Available options are: | ||
| * none: regular periodic calculation without isolated-system correction. | ||
| * makov-payne, m-p, mp: compute the Makov-Payne correction to the total energy and estimate a corrected vacuum level for eigenvalue alignment. This option is available only for cubic lattices (latname = sc, fcc, or bcc). | ||
|
|
||
| Theory: G. Makov and M. C. Payne, Phys. Rev. B 51, 4014 (1995). | ||
| default_value: none | ||
| unit: "" | ||
| availability: "" |
Collaborator
Author
|
Some PRs are making changes to docs without following the instructions after v3.9.0.24, so the documentation lags behind the source code. For example, #7384, #7354, #7280 update docs or add new params but did NOT update |
Collaborator
|
My practices for mixing_beta tuning: 0.4 -> 0.1 -> 0.025 is a general way |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Linked Issue
Fix #7125
What's changed?
mixing_beta&mixing_beta_magto suggest smaller parameter values for scf hard to converge, as in BUG: USPP hard to converge and yields incorrect magnetic moments #7125.