htmd.adaptive.util module

htmd.adaptive.util.getEpochSimIdx(data: MetricData, epoch: int) → ndarray

Return the simulation indexes in a MetricData object for a given epoch.

Parameters:

• data (MetricData) – Projected simulation data whose simlist is searched.

• epoch (int) – The epoch number to select.

Returns:

idx – Integer array of simulation indexes in data.simlist that belong to the requested epoch.

Return type:

np.ndarray

htmd.adaptive.util.getEpochTrajectoryDictionary(simlist: list) → dict

Build a mapping from epoch number to simulation indexes.

Parameters:

simlist (list) – A simulation list created using the simlist function.

Returns:

simepochs – A dictionary mapping epoch number (int) to a list of simulation indexes from simlist belonging to that epoch.

Return type:

dict

htmd.adaptive.util.getParentSimIdxFrame(data: MetricData, trajidx: int) → tuple

Find the parent simulation index and frame for a given trajectory.

For a simulation that was spawned from a parent trajectory, returns the index of the parent simulation within data.simlist and the frame index within that parent simulation.

Parameters:

• data (MetricData) – Projected simulation data containing the simlist to search.

• trajidx (int) – Index of the simulation in data.simlist for which to find the parent.

Returns:

• parentidx (int) – Index of the parent simulation in data.simlist.

• frameidx (int) – Frame index within the parent simulation corresponding to the spawn point.

htmd.adaptive.util.updatingMean(oldmean: float, oldcount: int, newdata: ndarray) → float

Compute a running mean incorporating new observations.

Parameters:

• oldmean (float) – The previously accumulated mean value.

• oldcount (int) – The number of observations that produced oldmean.

• newdata (np.ndarray) – Array of new observations to incorporate.

Returns:

mean – Updated mean over all observations (old and new combined).

Return type:

float