This is not available in Mplus. You will need to do this outside of Mplus.
Thomas Kirsh posted on Thursday, September 12, 2019 - 11:21 am
I have the same question and am wondering if this has been added recently. The goal of randomly splitting the data is to test the stability of the clustering by comparing the selected number of classes, the sizes of each class, and how often a subject is grouped in the same class. Does Mplus have this capability now or do I need to use R to do that?