Trajectory Load Helper

GitHub Link to Code.

MD trajectory loading strategies and utilities.

Handles the loading methodology for MD trajectories from files and directories. This class can be easily replaced or extended for different loading strategies. Supports nested directory structures and trajectory concatenation.

class mdxplain.trajectory.helper.process_helper.trajectory_load_helper.TrajectoryLoadHelper

Internal utility class for loading MD trajectories from files and directories.

Supports nested directory structures and trajectory concatenation.

static load_trajectories(data_input: List[Any] | str, concat: bool = False, stride: int = 1, selection: str | None = None, use_memmap: bool = False, chunk_size: int = 1000, cache_dir: str = './cache') → Dict[str, List[Any]]

Load trajectories from various input types.

Parameters

data_inputlist or str

List of trajectory objects or path to directory

concatbool, default=False

Whether to concatenate trajectories per system

strideint, default=1

Frame striding parameter

selectionstr, optional

MDTraj selection string to apply to each trajectory before concatenation

use_memmapbool, default=False

Whether to use memory-mapped DaskMDTrajectory for large files

chunk_sizeint, default=1000

Chunk size for DaskMDTrajectory (only used when use_memmap=True)

cache_dirstr, default=”./cache”

Cache directory for DaskMDTrajectory (only used when use_memmap=True)

Returns

dict

Dictionary with ‘trajectories’ and ‘names’ keys containing list of loaded trajectory objects and their corresponding names

Examples

>>> # Load from trajectory list
>>> trajs = [traj1, traj2, traj3]
>>> result = TrajectoryLoadHelper.load_trajectories(trajs, concat=True)
>>> print(f"Loaded {len(result['trajectories'])} trajectories")
>>> 
>>> # Load from directory
>>> result = TrajectoryLoadHelper.load_trajectories('../data', use_memmap=True)
>>> # Load with selection
>>> result = TrajectoryLoadHelper.load_trajectories('../data', selection='protein')