Based on my experience on Academia Stack Exchange, in a parallel universe there is a student complaining that their advisor made them rework everything the nigh

Based on my experience on Academia Stack Exchange, in a parallel universe there is a student complaining that their advisor made them rework everything the night before a presentation. Candidly, I probably would have told you to present it as is, and add a caveat to the last slide that this was work in progress (most internal presentations are assumed to be so) and you're still chasing down some problems in your code. The reason for this is two-fold: 1) It's the night before. Many students I have known and mentored aren't at the point in their career that they can "wing" a major presentation. It would be setting them up to fail in a way I couldn't shield them from (I can deal with changing results, but a bad presentation is largely on the student). 2) The quality of your checking is likely to be poor the night before. There's a number of times I've found an error as something was being prepped for presentation/poster printing/etc. and been convinced it changed everything, only to discover after 48 hours of thought and more checking that the difference in results was pretty negligible - especially in the sense of the qualitative take aways from a presentation. This, of course, may not have been your exchange with your PI. But I thought it was worthwhile that there are reasons not to have you change everything the night before that aren't the result of villainous fraud.