update documentation
This commit is contained in:
@ -67,6 +67,16 @@ resampled from a normal distribution:
|
|||||||
The algorithm then continues, proposing a new configuration of particles
|
The algorithm then continues, proposing a new configuration of particles
|
||||||
and velocities N integration steps later.
|
and velocities N integration steps later.
|
||||||
|
|
||||||
|
Typically, HMC is run with a larger timestep than molecular dynamics.
|
||||||
|
The timestep provides control over the acceptance ratio. A larger
|
||||||
|
timestep will lead to larger and more extreme MC moves which are
|
||||||
|
less likely to be accepted.
|
||||||
|
|
||||||
|
This fix is not designed to be used with anything but an NVE simulation.
|
||||||
|
Only atom-specific data are restored on MC move rejection, so anything which
|
||||||
|
adds or remove atoms, changes the box size, or has some external state not
|
||||||
|
dependent on atomic data will have undefined behavior.
|
||||||
|
|
||||||
----------
|
----------
|
||||||
|
|
||||||
The keyword/value options are used in the following ways.
|
The keyword/value options are used in the following ways.
|
||||||
|
|||||||
Reference in New Issue
Block a user