(1) Reduce the plugin code.
To make a more clean and maintainable code, the idea is to export all the Yambo-related functions and methods into yambopy, e.g.
- convergence analysis
- parsing of results from yambo (partially borrowed from yambopy)
- inputs generation (already massively borrowed from yambopy)
In this way the plugin will be simple to maintain, and we can focus on adding more complex work chain with additional features, e.g. yambopy-band interpolations (electronic and excitonic bands) ...
(2) Direct use of yambopy.
With this in mind, we have also to consider to add automatic error handling for yambopy errors (like indeed interpolation errors).
A clean way to do would be to create (restart) workchains to run yambopy commands, meaning that the plugin will be able to use yambopy also as remote executable, not only as local post processing set of modules. The plugin will then be an AiiDA plugin for both yambo and yambopy.
(1) Reduce the plugin code.
To make a more clean and maintainable code, the idea is to export all the Yambo-related functions and methods into yambopy, e.g.
In this way the plugin will be simple to maintain, and we can focus on adding more complex work chain with additional features, e.g. yambopy-band interpolations (electronic and excitonic bands) ...
(2) Direct use of yambopy.
With this in mind, we have also to consider to add automatic error handling for yambopy errors (like indeed interpolation errors).
A clean way to do would be to create (restart) workchains to run yambopy commands, meaning that the plugin will be able to use yambopy also as remote executable, not only as local post processing set of modules. The plugin will then be an AiiDA plugin for both yambo and yambopy.