batchtools doesn't use the specified seed as the actual seed for jobs but instead adds it to job.id. For my experiments I needed to fix the seed for all jobs to make result comparison fair and easier to do. I added fix.seed argument for both problems and jobs so there's at least an option to reuse the same seed for all jobs. If you ignore fix.seed arguments, I believe the package behaves the same as it used to. Thanks!
Coverage increased (+0.05%) to 87.53% when pulling f61554240fa7162a649318d43a60c73b410dcb11 on damirpolat:master into 03381348666d224178ae348fcbca8b6260abed43 on mllg:master.
batchtools doesn't use the specified seed as the actual seed for jobs but instead adds it to job.id. For my experiments I needed to fix the seed for all jobs to make result comparison fair and easier to do. I added fix.seed argument for both problems and jobs so there's at least an option to reuse the same seed for all jobs. If you ignore fix.seed arguments, I believe the package behaves the same as it used to. Thanks!