You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
SD seems to use a counter based noise source, as do the other thermostats. The counter value is derived from the simulation
time by calculating sim_time / time_step, which means that if the users resets the time, the noise is repeated. This differs from
the other thermostats, where the noise is derive from a monotonous counter which can not be reset. Also the SD version leads
to correlated noise if the time step is changed.
The text was updated successfully, but these errors were encountered:
Description of changes:
- Remove RNG correlation stemming from seed offsets (fixes#3585)
- seeds are now used as keys
- a monotonically increasing counter is used in each thermostat
- the only way to reset these counters is to create a new `System`
- Remove RNG correlation stemming from resetting `sim_time` or `time_step` during simulations with SD (fixes#3840)
- the SD thermostat now uses the same RNG interface as other thermostats
- Accelerate RNG unit tests (fixes#3573)
- they now take 2 seconds to run in coverage and sanitizer builds in CI
- Separate thermostats from integrators
- better separation of concerns
SD seems to use a counter based noise source, as do the other thermostats. The counter value is derived from the simulation
time by calculating
sim_time / time_step
, which means that if the users resets the time, the noise is repeated. This differs fromthe other thermostats, where the noise is derive from a monotonous counter which can not be reset. Also the SD version leads
to correlated noise if the time step is changed.
The text was updated successfully, but these errors were encountered: