Hey All,
In some cases, a group of structures is used in which not all required pseudopotentials are available in the specified pseudopotential family. This leads to that scenario: each time a new batch of calculations is submitted, the code iterates over all structures to identify the next valid one for submission.
A useful improvement ( I think ) would be a filtering step based on the pseudopotential family specified in the settings.yaml file. This filter would exclude any structures for which a complete set of pseudopotentials is not available.
This would avoid unnecessary iteration. What do you think ?