loads state of model and optimizer :dic: (empty) dictionary to fill with state information :datetime: date and time from run

log index values-histogram couple for specific item to tensorboard :item: string describing item (e.g. "training_loss","test_loss")

log index model-gradients-histogram couple for specific item to tensorboard :item: string describing model item (e.g. "encoder","disc

L2 loss on position.

compose ghost cell, but TODO: corner cell can not be duplicated twice with current cell type

mask_factor belongs between [0,1]

phi_node: (N, k) mesh_pos: (N, 2) edge_index: (2,E) bidirectional

transform dict to list, so pandas dataframe can handle it properly

saves dictionary - helpful to save the population state of an evolutionary optimization algorithm :dic: dictionary to store :

Splits trajectories into frames, and adds training noise.

'boolean type variable' for add_argument

'boolean type variable' for add_argument

returns delta t from last call of t_step()

Computes mesh edges from triangles.