parallel

Dask helper functions.

`build_gpu_cycler(workers_per_gpu=1, n_gpus=nr_gpus())`

Generator to cycle through GPU indices.

On machine with multiple GPUs and a computation that does not use a full GPU. This will yield GPU indices in a round-robin fashion.

If n_gpus is set to 0, it will yield 0 indefinitely.
If workers_per_gpu>0 and n_gpus=1, it will yield 0 indefinitely.
If workers_per_gpu=1 and n_gpus=2, it will yield 0, 1 indefinitely.
If workers_per_gpu=4 and n_gpus=2, it will yield 0, 1, 0, 1, 0, 1, 0, 1 indefinitely.

Configure the Dask scheduler by reusing existing or creating a new cluster.

When creating a local GPU cluster on a machine with multiple GPUs, it will start workers which each can only see a single GPU.

Parameters:

Name	Type	Description	Default
`scheduler_address`	`str \| Cluster \| None`	Address of the Dask scheduler to connect to, or None for local cluster.	required
`name`	`str`	Name for the Dask cluster.	required
`workers_per_gpu`	`int`	Number of workers per GPU. If > 0, a GPU cluster will be configured otherwise a CPU cluster.	`0`
`nproc`	`int`	Number of processes to use per worker for CPU support.	`1`

Raises:

Type	Description
`ImportError`	If GPU support is requested but pyopencl is not installed.
`ValueError`	If multiple GPUs are detected but the vendor is unsupported.

Returns:

Type	Description
`str \| Cluster`	A Dask Cluster instance or a string address for the scheduler.

The number of available GPUs on the system.

Returns:

Type	Description
`int`	Number of GPUs available

Worker function for running PowerFit on a single PDB file.

Parameters:

Name	Type	Description	Default
`pdb_file`	`Path`	Path to the PDB file to process	required
`density_map_target`	`Path`	Path to the density map file	required
`powerfit_run_root_dir`	`Path`	Root directory for PowerFit results	required
`options`	`PowerfitOptions`	PowerFit options	required