Summer release with many new features and bug-fixes

[sblisesivdin]

This is the fourth release of gpaw-tools in 2022. The most important feature added in this release is the successful implementation of different structure optimization minimizers. Users can choose QuasiNewton, GPMin, LFBGS, or FIRE minimizer for geometry optimization. Also, one big difference is the quick_optimization.py script. This script can now work just like gpawsolve.py and its name is now asapsolve.py. Therefore old script and folder are removed. You can find one example in examples folder. Also, there are many minor fixes and features added.

We kindly ask you to help us with your issues and pull requests. We have only one rule, try to keep it simple for broad-spectrum users.

Thank you for trying and helping to gpaw-tools to make it better.

Release notes

• optimize_latticeparam.py now can work for both lattice params a and c. Also draws 3D fig of energy-dependent latt_a - latt_c.
• quickoptimize.py works like gpawsolve.py now. Its name is now asapsolve.py.
• New default optimizer is QuasiNewton (BFGSLineSearch).
• New keyword Optimizer. Users can now choose QuasiNewton, GPMin, LFBGS, or FIRE minimizer for geometry optimization.
• gg.py includes all-new keywords.
• Grid point density or manual grid points for axis (LCAO only).
• Include new keywords for LBFGS geometry optimization Damping, Alpha, and Max_step.
• Geo_optim keyword for better optimization usage with filters (whichstrain).
• Examples are simplified. Most of the unnecessary keywords are removed.
• Proper logging for LCAO ground-state calculations
• Fix LCAO spinpol calculation (Thanks to Toma Susi).
• Include new keyword Mixer_type.
• Fix help description text width problem.
• Execution timing data of all calculations are saved to FILENAME-6-Result-Log-Timings.txt file.
• Instead of direct execution, all tasks are added to the task-spooler queue in do_all_examples.sh script.

---

This discussion was created from the release Summer release with many new features and bug-fixes.