An unfortunate typoo
Yes, the title was on purpose. No, I’m not thrilled at the moment. Sometimes you just need things to go smoothly, but life has other plans and yesterday I hit one hell of a stumbling block. The good news is I’ve caught it, but the bad news is there is now about 104833423x more work for me to do to fix the issue. No matter how careful you are, something is always going to get missed, yesterday was just a reminder that you can miss things even when you’re paying close attention. I may go as far as to say, especially when you’re paying close attention.
The story goes something like this, my Co-PI had a super cool experiment and asked if I wanted to do the analysis for it (here). I jumped on the chance like someone was going to snatch it up before I could. It was a super cool experiment, of course I wanted it! Well the due date for submission to the journal was technically last month, thankfully we got an extension because we are definitely not done yet! It’s been a mad rush to finish ever since and my Co-PI has informed me as of Sunday that he’s hoping to have this to them sometime next week.
No worries, I already did basically 100% of the analysis I just needed to write the results and the methods for the paper, a task which got me first authorship! Yep, this will be the second paper out of five that I’m going to be first author on. Second denoting the number of submissions I’ve already done, so only one paper in review right now (don’t remind me… so much writing!!), but things are moving quickly. Speaking of quickly you know how easy it is to make a mistake when you’re rushing through something?
Yesterday I had a bad feeling about part of the analysis I did. The data didn’t make sense and it confused me. I went back and checked the figure code I wrote, but there was nothing wrong there. Then I decided to pull the trigger and just look back through the super complicated analysis I had performed. All my inputs were good, everything checked out so I couldn’t figure out what the problem was with it. I was about to assume that maybe that’s just the data and I need to accept that it wouldn’t be perfect. After all people are complex things and anytime we do a human study we need to deal with that complexity, maybe this was just a part of that. Maybe our sample size was too small, or maybe the experiment wasn’t setup perfectly. Whatever the problem figured that I did my duty in checking my work and we’re reporting the data honestly… or that was what I was about to suggest to my Co-PI.
Then I saw the problem. The command I used to calculate the significance level was not exactly correct. The command should be: COMMAND( XX, YY, 2, 1), instead I had done COMMAND(XX, YY, 1, 2). Which is not a huge deal… except for the fact that the command did the wrong thing because I’m an idiot. Now all my fancy plots are wrong, everything I’ve written is wrong, and I spent the past FIVE (!!) hours yesterday going through and tediously fixing and updating everything so I had the correct output to make my plots from. Did I mention I’m sort of on a time crunch here? Yeah, the kind of time crunch that I really couldn’t spend a whole day debugging something, but ended up spending a whole day debugging anyway.
The good news is the data are correct now, the plots are mostly made, and the data still doesn’t make perfect sense, but it’s better. Some of the stuff we were looking for we didn’t find, some of the stuff we found we weren’t expecting, but that’s research. I’m sure we’re going to have to go through everything one more time to be sure anyway. A different paper I helped with (I’m only second author on that… wahh!!) we repeated the analysis like four different times by three different people. They all came back the same, but the point is we can get carried away a little. I’m assuming that even though I’m 99% sure I have the correct solution now, it will probably get checked a few more times by others.
As for the rest of the stuff that needs to happen today… well I’m behind schedule now which is never a good thing. On the bright side I caught a somewhat obvious error, so now I don’t have that looming over me. Still, I hate having so much work to do. I really wanted to take some time off this summer for myself, but unless I can get this all done the two weeks I wanted to take off before school starts again are just going to be wishful thinking. Speaking of which, I should get back to work if I want any hope of having some time to myself.