KratosMultiphysics
[Core] Adding clear to assign processes
Fixes #6291
This PR adds a clear method for the non-historical assign processes. It check the interval, if outside of the interval clears (set to zero) the variable. In order to avoid conflict between different processes with different intervals, the MARKER flag is considered
But the process doesn't have a clear method. How do you want to call this in a list of processes?
But the process doesn't have a clear method. How do you want to call this in a list of processes?
Like #6291 is still in discussion this PR may not be the final solution
But the process doesn't have a clear method. How do you want to call this in a list of processes?
Ok, I have not understand the point. the idea was to be clled only in these processes as these are the conflictive ones, but sure. It makes more sense as you say
blocking to prevent accidental merges on behalf of the @KratosMultiphysics/technical-committee
Ping?
As discussed in the comments, the solution of adding a Clear is not generic, and the current behaviour is the expected one. A user can set to zero the loads either by prescribing a new loading on the next interval or eventually by adding a load by table.
for this reason @KratosMultiphysics/technical-committee does not agree about this PR
In any case, would you mind if I do an independent PR in order to define a virtual Clear function in the base Process?, many processes define a Clear function and every-time that this process is exposed to python the Clear must be re-declared
As discussed in the comments, the solution of adding a Clear is not generic, and the current behaviour is the expected one. A user can set to zero the loads either by prescribing a new loading on the next interval or eventually by adding a load by table. for this reason @KratosMultiphysics/technical-committee does not agree about this PR
In any case, would you mind if I do an independent PR in order to define a virtual Clear function in the base Process?, many processes define a Clear function and every-time that this process is exposed to python the Clear must be re-declared
@KratosMultiphysics/technical-committee strongly agrees with this. The addition of the Clear should be straightforward.
As discussed in the comments, the solution of adding a Clear is not generic, and the current behaviour is the expected one. A user can set to zero the loads either by prescribing a new loading on the next interval or eventually by adding a load by table. for this reason @KratosMultiphysics/technical-committee does not agree about this PR
In any case, would you mind if I do an independent PR in order to define a virtual Clear function in the base Process?, many processes define a Clear function and every-time that this process is exposed to python the Clear must be re-declared
@KratosMultiphysics/technical-committee strongly agrees with this. The addition of the
Clearshould be straightforward.
It is done, I added the Clear in the Base Process and override the existing one. Also added some Clear functions to important assign processes
You misunderstood, we meant a separated PR with only adding the Clear in the baseclass and fixing the overrides
As discussed in the comments, the solution of adding a Clear is not generic, and the current behaviour is the expected one. A user can set to zero the loads either by prescribing a new loading on the next interval or eventually by adding a load by table.
for this reason @KratosMultiphysics/technical-committee does not agree about this PR
see comment above
As discussed in the comments, the solution of adding a Clear is not generic, and the current behaviour is the expected one. A user can set to zero the loads either by prescribing a new loading on the next interval or eventually by adding a load by table. for this reason @KratosMultiphysics/technical-committee does not agree about this PR
see comment above
Therefore do I close this, or rename the method?
As discussed in the comments, the solution of adding a Clear is not generic, and the current behaviour is the expected one. A user can set to zero the loads either by prescribing a new loading on the next interval or eventually by adding a load by table. for this reason @KratosMultiphysics/technical-committee does not agree about this PR
see comment above
Therefore do I close this, or rename the method?
#6291 is not solved yet
As discussed in the comments, the solution of adding a Clear is not generic, and the current behaviour is the expected one. A user can set to zero the loads either by prescribing a new loading on the next interval or eventually by adding a load by table. for this reason @KratosMultiphysics/technical-committee does not agree about this PR
see comment above
Solved
As discussed in the comments, the solution of adding a Clear is not generic, and the current behaviour is the expected one. A user can set to zero the loads either by prescribing a new loading on the next interval or eventually by adding a load by table. for this reason @KratosMultiphysics/technical-committee does not agree about this PR
see comment above
Solved
Solved long time ago
The @KratosMultiphysics/technical-committee discussed this and agrees to add the method Clear to the baseclass of Process but it does NOT agree with the other implementations of this PR
my personal suggestion: Open a new PR where you add the Clear method to Process and close this one
The @KratosMultiphysics/technical-committee discussed this and agrees to add the method
Clearto the baseclass ofProcessbut it does NOT agree with the other implementations of this PRmy personal suggestion: Open a new PR where you add the
Clearmethod toProcessand close this one
What is the problem with the implementation?
What is the problem with the implementation?
This:
As discussed in the comments, the solution of adding a Clear is not generic, and the current behaviour is the expected one. A user can set to zero the loads either by prescribing a new loading on the next interval or eventually by adding a load by table.
for this reason @KratosMultiphysics/technical-committee does not agree about this PR
As discussed in the comments, the solution of adding a Clear is not generic, and the current behaviour is the expected one. A user can set to zero the loads either by prescribing a new loading on the next interval or eventually by adding a load by table. for this reason @KratosMultiphysics/technical-committee does not agree about this PR
Ok, but this PR was done in order to solve an issue. Of course, it changes behaviour, that's the point of the PR. The problem is that it requires explicitly to the user to clear the load after is applied, when it is supposed to disappear once the interval is finished
well the @KratosMultiphysics/technical-committee does NOT agree with this change. What might be expected for Structures is exactly the opposite for fluids! I don't know what to say otherwise :sweat_smile:
well the @KratosMultiphysics/technical-committee does NOT agree with this change. What might be expected for Structures is exactly the opposite for fluids!
But I don't understand, if you want to set some value until the end you can set an interval [something, "End"]. Can I discuss this with @KratosMultiphysics/technical-committee ?
What is the status of this¿? https://github.com/orgs/KratosMultiphysics/teams/technical-committee