-
-
Notifications
You must be signed in to change notification settings - Fork 12
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Handle crashing micro simulations #85
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Good first steps, but we need to do a bit more work before merging this. Please look at the comments below. A general comment is writing comments in the code such that it is readable. I will do another round of reviewing soon.
The latest commit adds interpolation. Besides lacking extrapolation, scipy's griddata module also cannot interpolate on linearly dependent data, thus data on the dummy problems cannot be interpolated. more thought must be put into the issue, to interpolate and extrapolate all cases of simulation crashes, and implementing a custom solution for interpolation is most likely required. |
Further discussions led to the idea that implementing an inverse distance weighing would make sense for the interpolation. This would be done with a k-nearest-neighbor method, where the user can set k. |
40c1f95
to
e6c3baf
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Good effort already 👍 my suggestions are mainly about code styling. Let us test this functionality on a real example and check if crashes are caught.
5ea8a70
to
1dc33ce
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We are getting closer to merging this 😁 some more comments on my side. The logic and overall code looks good now.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Good to go 👍 please remember to squash and merge.
This is an attempt to solve issue #74.
This cannot be a permanent solution to the handling of crashing simulation. If a simulation crashes during the first time step, it is replaced with data from another random simulation. In the long run, I don't see any other solution than an interpolation (in cases where only one simulation is given and it cannot be replaced by one of different complexity).
Moreover, it has not been tested with real simulations, in parallel, or using spack.