inconsistent result pytraj vs cpptraj when combining autoimage with rmsfit about pytraj HOT 5 CLOSED

this does not happen with Trajectory (mutable) class.

In [13]: traj_mut = traj[:]
In [15]: traj_mut.autoimage()
In [16]: traj_mut.rmsfit(mask='@CA,N,C')
In [17]: rmsd(saved_traj.xyz, traj_mut.xyz)
Out[17]: 5.75140630497023e-07

from pytraj.

Order matters here. Once you RMS-fit, imaging doesn't work. Maybe your first example is RMS-fitting, then autoimaging?

from pytraj.

@swails
It seems that when iterating frame for TrajectoryIterator, I did the imaging first.
https://github.com/hainm/pytraj/blob/master/pytraj/core/frameiter.py#L64

I always have trouble debugging things related to rms or rmsd because it changes coordinates of frame and reference frame. I need to double-check above code, must be mistaken somewhere.

from pytraj.

oops, I just updated above link.

from pytraj.

@swails
now I know why, I did not do autoimage for the reference structure before do fitting. :D

from pytraj.