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
I am running Galacticus on N-body merger trees constructed with the code Consistent Trees. When the trees are constructed with this code, hosts and their subhalos can apparently belong to different trees. This causes an error in Galacticus:
Fatal error in Build_Descendent_Pointers():
failed to find host node: 42283 of 42279
I thought I might solve the problem by changing the argument in this call (in merger_trees.file_maker.simple.F90):
call mergerTrees%setSelfContained(.true.)
from ".true." to ".false.". However, when I do this, I get the error:
Fatal error in Merger_Tree_Read_Initialize():
only self-contained trees are supported
Is this indeed the case, (or can this be due to a conflict with some other parameter or setting, for example)? If so, do you have any suggestions for how to solve the problem? For now, I have created a small script that goes through the tree and changes the tree indices such that hosts and subhalos have the same tree index. However, this is rather time-consuming for large trees.
The text was updated successfully, but these errors were encountered:
Original comment by Andrew Benson (Bitbucket: abensonca, GitHub: abensonca).
In these cases you have two options. The correct thing to do is to take all interconnected trees - referred to as a "forest" in Rockstar/ConsistentTrees - and include all halos from these trees as a single forest in the merger tree file that you construct for Galacticus. ConsistentTrees outputs forest information which allows you to do this. Galacticus will then correctly handle these cases.
which tells Galacticus to ignore missing hosts. This will allow you to run the trees through Galacticus, but the results you get will not be consistent with the true behavior of the N-body simulation.
Original report by Io Odderskov (Bitbucket: ioodderskov, GitHub: ioodderskov).
Hi,
I am running Galacticus on N-body merger trees constructed with the code Consistent Trees. When the trees are constructed with this code, hosts and their subhalos can apparently belong to different trees. This causes an error in Galacticus:
I thought I might solve the problem by changing the argument in this call (in merger_trees.file_maker.simple.F90):
from ".true." to ".false.". However, when I do this, I get the error:
Is this indeed the case, (or can this be due to a conflict with some other parameter or setting, for example)? If so, do you have any suggestions for how to solve the problem? For now, I have created a small script that goes through the tree and changes the tree indices such that hosts and subhalos have the same tree index. However, this is rather time-consuming for large trees.
The text was updated successfully, but these errors were encountered: