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
There is some basic considerations towards this, but a more thorough audit needs to be done to make sure that this is supported in a first class way.
Currently, if the number of walkers increases a new "trajectory" is added and in the trajectory group metadata a field is added noting which cycle it was created.
The text was updated successfully, but these errors were encountered:
The first approach was to hack this into the "sparse fields" feature.
Options to make all fields sparse for variable number simulations was added (PR forthcoming).
This mostly works but is broken wrt mandatory "fields" i.e. weights. The proper solution then is support a dataset under each trajectory of _cycle_idxs and then _sparse_idxs for each field is relative to that. This allows for them both to coexist, which was a problem with the sparse fields for variable number of walkers hack.
There is some basic considerations towards this, but a more thorough audit needs to be done to make sure that this is supported in a first class way.
Currently, if the number of walkers increases a new "trajectory" is added and in the trajectory group metadata a field is added noting which cycle it was created.
The text was updated successfully, but these errors were encountered: