COMBO
COMBO copied to clipboard
QUVA-Lab
What is the role of the sorted_partition variable?
Hi,
I have a question about the sorted_partition variable. I have noticed that it is used to compute edge_mat_list in posterior_sampling and by the group_input function. In the first case, edge_mat_list is always equal to the input adj_mat_list, and in the second case, the grouped inputs appear to be identical to the input arguments to the group_input function. So in both cases it does not appear as if the two functions are doing anything, hinting that the sorted_partition variable might be redundant.
I wanted to ask you what exactly is the role of this variable and what is its use? I have also noticed that in the main.py file it is always initialized to sorted_partition = [[m] for m in range(eval_inputs.size(1))] irrespective of the task, and I imagine that this is why the two above mentioned functions simply return their inputs. When would you initialize sorted_partition another way in practice?
Hi.
You can ignore that partition. It was intended for some extension of COMBO I tried. I haven't cleaned that up .
On Sat, Mar 26, 2022 at 1:23 AM KamilDre @.***> wrote:
Hi,
I have a question about the sorted_partition variable. I have noticed that it is used to compute edge_mat_list in posterior_sampling and by the group_input function. In the first case, edge_mat_list is always equal to the input adj_mat_list, and in the second case, the grouped inputs appear to be identical to the input arguments to the group_input function. So in both cases it does not appear as if the two functions are doing anything, hinting that the sorted_partition variable might be redundant.
I wanted to ask you what exactly is the role of this variable and what is its use? I have also noticed that in the main.py file it is always initialized to sorted_partition = [[m] for m in range(eval_inputs.size(1))] irrespective of the task, and I imagine that this is why the two above mentioned functions simply return their inputs. When would you initialize sorted_partition another way in practice?
— Reply to this email directly, view it on GitHub https://github.com/QUVA-Lab/COMBO/issues/8, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABJKCME2ECDFXAW34TKIOM3VBXSBZANCNFSM5RUTWAPQ . You are receiving this because you are subscribed to this thread.Message ID: @.***>
