Our current CLIs take single positions as input, but for large datasets this approach requires the user to manage SLURM, HPC resources, and multiple bash/python scripts.
Here we would like to prototype a revised CLI that has the following features:
handles resource requests for the user. For example, a phase reconstruction should always require a fixed multiple of the input size of RAM...no need for the user to calculate this.
talonchandler
commented
3 days ago
Our current CLIs take single positions as input, but for large datasets this approach requires the user to manage SLURM, HPC resources, and multiple bash/python scripts.
Here we would like to prototype a revised CLI that has the following features: