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
Is there any merit to always saving when finalize() is called? (dropping the if conditional)
Yes this would potentially throw off a regular save interval if the run is not due for a save. But, this would ensure that the 'final' state of the model is captured (might be nice in the case of a run with save_dt = 150, dt = 15 and a final time of 295 as an example?). Otherwise a run could just go on and run many iterations, depending on the configuration, that are ultimately not saved.
Alternatively I could see the benefits of outputting a warning to the user telling them that X number of timesteps or amount of time will be run but not saved. Or using the save_dt and dt information to cut a run short and avoid wasted iterations... These concerns are not really pertinent to this PR, just something that occurred to me.
Is there any merit to always saving when
finalize()
is called? (dropping theif
conditional)Yes this would potentially throw off a regular save interval if the run is not due for a save. But, this would ensure that the 'final' state of the model is captured (might be nice in the case of a run with
save_dt = 150
,dt = 15
and a final time of295
as an example?). Otherwise a run could just go on and run many iterations, depending on the configuration, that are ultimately not saved.Alternatively I could see the benefits of outputting a warning to the user telling them that
X
number of timesteps or amount of time will be run but not saved. Or using thesave_dt
anddt
information to cut a run short and avoid wasted iterations... These concerns are not really pertinent to this PR, just something that occurred to me.Originally posted by @elbeejay in #78
In reference to first
if
conditional infinalize()
The text was updated successfully, but these errors were encountered: