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
An alternative would be to use Serialization.serialize or JLD2.jl.
However, we should also think of a mechanism that would allow us to restart longer calculations that ran out of time, crashed due to power cuts or other emergency shutdowns, etc. For this we would need a robust way to save the current state of a long calculation on regular time intervals, and restarting from such data.
A related issue is the saving of metadata for a MC calculation to file. This would be useful for restarting calculations, but could also be useful for automated data analysis and plotting scripts, as the dataframes produced by lomc! currently lack information about most of the calculation parameters.
The text was updated successfully, but these errors were encountered:
Rimu is providing the functions
save_dvec
andload_dvec
for savingDVec
s to disk. Apparently the packageBSON.jl
on newer Julia versions ≥ 1.8. Consequently these function should no longer be used. See https://github.com/joachimbrand/Rimu.jl/blob/8f91d7d2f7b7e36ac349e79ad410ece7404489d6/test/runtests.jl#L93An alternative would be to use
Serialization.serialize
orJLD2.jl
.However, we should also think of a mechanism that would allow us to restart longer calculations that ran out of time, crashed due to power cuts or other emergency shutdowns, etc. For this we would need a robust way to save the current state of a long calculation on regular time intervals, and restarting from such data.
A related issue is the saving of metadata for a MC calculation to file. This would be useful for restarting calculations, but could also be useful for automated data analysis and plotting scripts, as the dataframes produced by
lomc!
currently lack information about most of the calculation parameters.The text was updated successfully, but these errors were encountered: