The big presentation
It has been a day friends. Yesterday I got an email from my main-PI asking me about a funding proposal I’ve been working on and when I responded I asked if we were having our lab meeting this week since the last two have been cancelled. He said yes and he mentioned that today would be the day I presented the work I’ve been doing with my project. Thankfully I was ready for it, but it was pretty stressful and there are some changes that need to be made. Overall the lab seemed pretty impressed by the presentation and the work I’ve been doing.
For those who have no clue what’s going on, over a year ago now I did an experiment (one that I wasn’t particularly thrilled about (more). It was a tough experiment, the dataset was hard to work with, and the whole thing was completely experimental. I mean that in the sense that the data had never been collected before, so we were doing a first, so there was nothing even remotely close to compare with.
It’s been slow, like really slow. I had basically no support here and I was doing everything by hand, for the first time, it was not fun. Still, I’ve been hard at work and we’ve finally made it to the end of the data processing! Today was the last milestone I needed to clear and it was the one where I got to show off my fancy figures (here) that I thought were going to impress everyone. I also struggled a lot with that (here), but in the end I completed the project.
Which brings us to today. I was nervous going into the presentation because a lot of this (okay all of this) was custom written code to do all the stuff that needed to happen. We have software to do all the work for us, but I never really learned how to use it and by the time it came necessary to use the software I had already done most of the process in a different way that I couldn’t quite get to work in the premade programs that we needed to use. Moreover, some of those programs hid a lot of the steps they did so it was hard to follow along and figure out what the software was actually doing. In fact, one of the other people in the lab discovered that the program we were running wasn’t doing the thing we thought.
I ended up writing a lot of code for myself because it was easier to keep track of what was happening than it was to trust the program. Plus I needed stuff in-between that the program wouldn’t readily give me, so even if I took the time to learn the code, reformat my data to be the correct way, and did all the stuff correctly, I still couldn’t get the data back that I wanted. The catch was I needed to make sure my code was correct. So there was a lot of verifying and double checking that the stuff I was doing was correct, but I also had serious doubts about my work.
Interestingly more than 90% of my presentation was covering the workflow to my fancy figures. I went in excruciatingly painful detail about the steps I used because I wanted to make it clear I did this the best way possible. I even went as far as adding in the dimensions of the data at each step since I was doing different calculations with it and breaking it up into different chunks. Which meant I needed to read off several strings of numbers over and over while remembering what each of those numbers actually accounts for so for example I had a 5 x 5 x 87 x 28 x 8 matrix in one of my steps and needed to explain what each of those dimensions meant.
Then I FINALLY got to show off my figure and it was a pretty big hit! Everyone enjoyed it and I got some great suggestions about little changes I could make to improve them. I also got some good feedback about additional things I could add to the figures themselves. I say figures, but they were more videos, it’s hard to explain, but hey as soon as we publish this I will be more than happy to share what I did because it’s so cool!
Anyway my main-PI (soon to be my only PI…) seemed to be pretty impressed with the work and asked me a bunch of questions about how quickly we could implement something like this for real-time data processing for presentations and outreach, which since I wrote the code and it’s super “lean” without all the extra stuff squeezed into the software it runs very quickly and could in theory be implemented in real-time. So I may have just created a new tool that the lab can use for outreach and what not.
Overall it was a good day, a busy day for sure, but good. Tomorrow I’ve got experiments and like 50 things all due this week, but I’ll take the win when I can get it. For once I feel pretty good about my place in the lab, which has been a rare occurrence. It’s taken me a lot of time to get used to the way the lab does things, but I’m glad that I’m in a good spot now. Now I just need to get everything else done and this paper written, but I am hopeful… for once.