>>212171826 (OP)When I was working on my dissertation, part of my work was computational stuff and I realized about two weeks before my defense that something didn't look right about my results. I spent a solid four days pouring through my code line by line, writing down individual data points on post its to try and correlate all the inputs and outputs and figure out what went wrong, trying to re-run the code to see if I could replicate my previous results, and all the while I was sleeping maybe 2-3 hours a night from all the anxiety and my whole apartment looked like pic related.
I finally got everything sorted, realized all my data was fine and that I'd just transposed two numbers in one of the vectors I was using to project the data for all my figures and that fixed everything. Then I drank a bunch of scotch, passed out, and slept for sixteen hours straight.