Center

GeometricCenterTask

Bases: Task[NDArray[float64], NDArray[float64]]

A Task to compute the geometric center for each structure (microstate) within a given chunk of atomic coordinates.

This task is designed to work with chunks of coordinates, making it suitable for large datasets and parallel processing. It expects input as a NumPy array where each row represents a microstate and contains the flattened coordinates of all atoms for that microstate, or a 3D array (n_microstates, n_atoms, 3).

The geometric center is calculated as the mean of the x, y, and z coordinates for all atoms in each structure.

npt.NDArray[np.float64]

A NumPy array representing a chunk of atomic coordinates. Expected shape: (n_microstates, n_atoms, 3)

npt.NDArray[np.float64]

A NumPy array where each row is the geometric center (x, y, z) for a corresponding microstate in the input chunk. Expected shape: (n_microstates_in_chunk, 3)

Examples:

```python
from raygent import TaskManager
from atomea.tasks.geom import GeometricCenterTask

# Instantiate the TaskManager with the new task
center_task_manager = TaskManager(
    task_class=GeometricCenterTask,
    use_ray=False,
)

# Submit tasks using the .iter() method from Ensemble.coordinates
# The chunk_size here determines how many microstates are processed at once
# by the `process_items` method of GeometricCenterTask.
center_task_manager.submit_tasks(
    items=ensemble.coordinates.iter(run_id="run_01", chunk_size=10),
    at_once=True,  # Important: tells Task.run to call process_items
)

# Retrieve results
geometric_centers = center_task_manager.get_results()
```

do(batch, *args, **kwargs)

Computes the geometric center for each structure in the input chunk.

PARAMETER	DESCRIPTION
items	Atomic coordinates of one or more microstates.
**kwargs	Additional keyword arguments (not used in this specific task). TYPE: object DEFAULT: {}

RETURNS	DESCRIPTION
Float64	A list of NumPy arrays, where each array is the geometric center (x, y, z)
Float64	for a corresponding microstate. Each inner array will have shape (3,).