The Altair Community is migrating to a new platform to provide a better experience for you. In preparation for the migration, the Altair Community is on read-only mode from October 28 - November 6, 2024. Technical support via cases will continue to work as is. For any urgent requests from Students/Faculty members, please submit the form linked here
Monitor progress of running process?
Is there a way to monitor the progress of a running process while it's "inside" a node?
For example, I have built an operator tree that contains a NeuralNet node. Evaluating that node takes the bulk of the processing time. I know the status bar at the bottom of the window indicates what node it is in, and how it has been running, but what I'd really like to is what percentage of the processing for the node is completed, or much longer it has to go. In this case, maybe how many training cycles have been completed. That would allow to estimate how long the process still has until it completes.
This specific case is a NeuralNet node, but I suspect it applies to other long-running nodes as well. Is this possible?
Thanks,
Keith
For example, I have built an operator tree that contains a NeuralNet node. Evaluating that node takes the bulk of the processing time. I know the status bar at the bottom of the window indicates what node it is in, and how it has been running, but what I'd really like to is what percentage of the processing for the node is completed, or much longer it has to go. In this case, maybe how many training cycles have been completed. That would allow to estimate how long the process still has until it completes.
This specific case is a NeuralNet node, but I suspect it applies to other long-running nodes as well. Is this possible?
Thanks,
Keith
0
Answers
there is unfortunately no easy way to do this in a generic way. Although it is not (yet) supported, here are some thoughts and ideas:
Way 1:
- one could add the ProcessLog operator as inner operator of a monitorable (<-- does this word exist?) operator
- the outer operator could provide the values which should be monitored
Although this would fit nicely to the process logging like it is supported right now, it would mean that monitorable must be operator chains instead of operators which would clearly a drawback and would probably not ease the process design.
Way 2:
- one could add a new operator "OperatorLog" before the operator which should be monitored
- this operator is used similar to a ProcessLog but adds a sort of listener to the monitored operator which is called each time new values arrive
I would probably prefer Way 2 if I would implement this type of inner-operator-monitoring.
Cheers,
Ingo