Great, thank you for the info. For additional context, I am creating an environment on codeocean that I can refer to in a journal article to reproduce an analysis. It sounds like it isn’t as straightforward as using random seeds to get identical results on there and on my local machine.