Config script #149
Config script #149
Conversation
…ocs contain volumetric data
Currently these collections are hard to search for specific structures because they create their own collections. These changes make it somewhat easier to find calculations within collections.
…ibution - Update docs for GibbsAnalysisToDb and the Gibbs wfs - Update options for preset gibbs workflow - Update base gibbs workflow - Update GibbsAnalysisToDb task - Fix updating metadata in GibbsAnalysisToDb task
There was a problem hiding this comment.
I like this effort and think having some baked in setup is really important.
I'm having some trouble understanding the utility of this. Are you still working on it?
It seems like it provides more or less the same defaults that you could get if you copied the setup files directly from Fireworks to your destination path.
I have something similar to this here, but hacked as a shell script. I've been meaning to move to Python for awhile and I think what you have here is similar to what I was envisioning.
Ideally, I'd like to see something that can walk users through a little better, especially because most users, at least in my experience have non-default files: not localhost mongo, path considerations, queue systems (EDIT: I see now that you have done this), etc. and neither of our solutions get very far past configuring files by hand.
I'm thinking along the lines of the interactive setup script for FWs launchpad or by trying to auto detect the queue system and letting the user choose from a list if we can't auto detect SLURM vs. PBS, etc.
I'd be happy to help write code or generate more ideas.
|
As another side comment, @computron has suggested (and the documentation reflects) using rapidfire instead of singleshot in the queueadapters. We should provide the option for users to use a reservation + singleshot approach, but IMO the default should be consistent with our docs to launch with rapidfire within the walltime. |
- Modify the EOS tutorial to use a much improved Gibbs free energy workflow example - Update old references to Si in running workflows Closes #124
|
I suggest looking at pymatgen-db's |
|
@bocklund I agree. I've coded this up for the use cases I've experienced in on-boarding undergrads to using atomate and put the pull request up to get feedback. I think a fully interactive mode would be a good idea, but shouldn't take away from a simple noninteractive mode. Let me add that in today. |
…n probably revisit when matminer is more mature
Small change to VaspDrone to pick up volumetric data metadata
Add anharmonic support to Gibbs free energy workflow
Atwf update
|
temp closing to clean up the pull request |
A simple config script for atomate.