Currently, the random seed is automatically shuffled when a job is cloned, or when a new job is created, unless it is explicitly altered.
I wonder if it would be possible to optionally (via a preference?) change this behavior, such that when a job is cloned, the random seen is maintained.
The rationale is as follows - when cloning a job, I am generally tweaking parameters - looking for the effect of changing a parameter on the outcome. But if the random seed is changed, I am looking at both the effect of the random seed and the parameter change. So generally when cloning jobs, I would rather the seed was cloned too. For ab initio in particular, the random seed can really make a difference, but I need to remember to manually fix it when cloning jobs.
Sorry for the delayed response! If you clone the initial job and change the seed to the same as the original job (eg. making it a non-default param), then the seed will be preserved when cloning the cloned job. Do you think that would work for your use case?
Hi Kye - thanks for the reply! That is what I do - I was just suggesting that the current defaults may not be optimal when screening parameters, and I’m not sure how many users pay attention to the random seed. I suspect that mostly when users are cloning jobs, they are doing to to test the effect of altering parameters, in which case fixing the seed would make sense, but probably I am overlooking something
For what it’s worth, I have the same request as Oli. The solution you proposed is the one I usually use as well, but it introduces an extra cloning step since I want all the clones to match an original with some set of default parameters, not the cloned descendant I’m making changes to.
