Networks and Transitions

Networks are an easy way to set up ensembles for path sampling calculations, and to retain a context for those ensembles when analyzing the results of a simulation. Networks are made up of multiple transitions, where each transition represents a directed \(A\to B\) reaction.

Many path sampling methods (transition interface sampling as one example) require sampling multiple path ensembles, and then results from those path ensembles are combined in analysis. As such, the path ensemble itself is not enough: you need both the path ensemble and its context. Networks provide that context.

Networks are so named because they allow the study of complicated transition networks, instead of just single transitions. It is at this level that we encounted ideas like multiple state path sampling methods and multiple interface set transition interface sampling.

A longer discussion of networks, and the associated concept of transitions, can be found in the file Transitions and Networks.

Abstract network class


Subclasses of TransitionNetwork are the main way to set up calculations

TPS networks

TPSNetwork(initial_states, final_states[, ...])

Class for flexible pathlength TPS networks (2-state or multiple state).

FixedLengthTPSNetwork(initial_states, ...[, ...])

Class for fixed pathlength TPS networks (2-states or multiple states).

TIS networks

MSTISNetwork(trans_info[, ms_outers])

Multiple state transition interface sampling network.

MISTISNetwork(trans_info[, ms_outers, ...])

Multiple interface set TIS network.


In general, users don’t need to directly create a Transition – the network acts as a factory that creates them.

Transition(stateA, stateB)

Describes (in general) a transition between two states.

TPSTransition(stateA, stateB[, name])

Transition using TPS ensembles

FixedLengthTPSTransition(stateA, stateB, length)

Transition using fixed length TPS ensembles

TISTransition(stateA, stateB, interfaces[, ...])

Transition using TIS ensembles.