I thought that it could be a problem that lefse may not reproduce the result. Here, it used python wrapper to call R function. Although python set seed is here but not for R at the Linear Discriminant Analysis step. I add r_objects.r("set.seed(123)"). But the replication is still different. I don't know why, but could you help me to clarify whether your team preferred to have random results or your team missed this function. If so, please help to fix this issue. Thank you.
