Repository navigation
Feature/parmoo - #58
Conversation
Initial draft of sdk
Work in progress
Fixed several bugs causing images and code blocks not to render. Added a few additional details to help new users get started.
parmoo readme patch
Added reference for the example problem and linked files.
|
Atting @thchang for awareness |
asemposki
left a comment
There was a problem hiding this comment.
Finished reviewing so far. Had one part of the README.md commands that might need to be changed, and one error importing the dash module which I do not appear to have. Should that be installed with parmoo in the requirements file?
mosesyhc
left a comment
There was a problem hiding this comment.
Reviewed parmoo in BAND.
I have observed similar behavior about installation and package dependencies as @asemposki, added my observations in the comments, and suggested certain changes. See comments in README.md.
I included a comment for a minor typo.
I verified the BAND SDK compliance, except for one. See comment under BAND SDK Mandatory Policy No.2.
…y overlaps with the basic parmoo installation
odell
left a comment
There was a problem hiding this comment.
- Add kaleido and dash to REQUIREMENTS.txt.
- Change command in README.md (bandframework) to:
git clone https://github.com/parmoo/parmoo-solver-farm
Band README revisions
|
Thanks @odell @mosesyhc @asemposki ! In addition to clarifying the README, the corresponding REQUIREMENTS.txt file in the parmoo-solver-farm/fayans-model-calibration-2022 repository has been updated to better pull in dependencies. |
asemposki
left a comment
There was a problem hiding this comment.
Made a note here about the difference in the plot I get vs the plot I'm expected to see. Not sure if they have to be the exact same. No code errors came up while running, and the REQUIREMENTS.txt file now contains all of the packages that I did not have previously installed, so everything looks good otherwise.
If it is OK to have the plots not be the exact same result, then consider this approval of the inclusion of parmoo in BAND.
mosesyhc
left a comment
There was a problem hiding this comment.
The updates have fixed the previous issues.
|
Thanks all |
This PR for the inclusion of ParMOO in the BAND.
Note that rather than being included as a submodule or have the source code in bandframework, the approach taken was to instead include sufficient documentation for the quick installation of ParMOO as well as how to obtain an example problem of ParMOO being run for a multi-objective calibration of (surrogate of) the Fayans energy density functional.
We are asking for reviewers to:
Thank you!