mdtraj.iterload

mdtraj.iterload(filename, chunk=100, **kwargs)[source]

An iterator over a trajectory from one or more files on disk, in fragments

This may be more memory efficient than loading an entire trajectory at once

Parameters:
  • filename (path-like) – Path to the trajectory file on disk

  • chunk (int) – Number of frames to load at once from disk per iteration. If 0, load all.

  • top ({str, Trajectory, Topology}) – Most trajectory formats do not contain topology information. Pass in either the path to a RCSB PDB file, a trajectory, or a topology to supply this information. This option is not required for the .h5, .lh5, and .pdb formats, which already contain topology information.

  • stride (int, default=None) – Only read every stride-th frame.

  • atom_indices (array_like, optional) – If not none, then read only a subset of the atoms coordinates from the file. This may be slightly slower than the standard read because it requires an extra copy, but will save memory.

  • skip (int, default=0) – Skip first n frames.

See also

load, load_frame

Examples

>>> import mdtraj as md
>>> for chunk in md.iterload('output.xtc', top='topology.pdb')
...    print chunk
<mdtraj.Trajectory with 100 frames, 423 atoms at 0x110740a90>
<mdtraj.Trajectory with 100 frames, 423 atoms at 0x110740a90>
<mdtraj.Trajectory with 100 frames, 423 atoms at 0x110740a90>
<mdtraj.Trajectory with 100 frames, 423 atoms at 0x110740a90>
<mdtraj.Trajectory with 100 frames, 423 atoms at 0x110740a90>