2

I am about to finish an undergraduate degree, with a good enough GPA that even if I got C's in all of my current classes, my final overall cumulative GPA would remain above 3.6.

Assuming I graduate with all the whistles in my diploma it would be:

Honours Bachelors of science in computer science with distinction and a minor in pure mathematics (quite pompous).

In order to complete all of those whistles I had to do supervised research for two semesters. Being a naive undergraduate, I knocked on proffessor's doors, asking for a supervisor, and one of them, very generously agreed. He suggested a problem I thought was interesting and so I started the research. I originally started throwing analysis at the problem, and started trying to solve some really complex differential equations (which kept me motivated). However I did not make much progress, so after my supervisor marked that I needed to get something, I fully changed my approach and started using some geometric approximations to solve the issue. This generated some results, but was too simple. I'd like to say that my results were inversely proportional to how much effort I was putting in the research.

Time passed, I had multiple potential solutions and a way to evaluate them. On the last week before the final submission I realised there was a bug in my code base that was destroying the results (they were mathematically impossible). I tried to find the error, but the best I could do was identify an area where a mistake could have happened.

Finally I got my evaluation on the course after presenting it. The feedback said the methods were a bit naive and could be improved on, and it was dissapointing that I didn't get numerical results (I had visual results, not numerical), however since I was not a seasoned researcher they decided to be lenient and gave me an A anyway.

I applied to the grad program at my university, and asked the same proffessor to be my supervisor for the grad program. I am in desperate need of financial support until the grad program starts, and so my supervisor said I could do summer research for him and in the meanwhile we could discuss what the grad project could be. What he wants me to do is finish my previous work and get a publication.

The issue is that I legitimately feel that dinking bleach would be more motivating (I am using this metaphor to convey just how much anxiety this is giving me). This research really does not appeal to me in any way or form, and the interesting part is over, all that is missing is finding the bug, which could take more than two weeks to solve and implementing a couple of modifications to make my naive methods less naive if things go well, or complete failure to finish it in the worst case.

In addition to that I have been doing independent research simultaneously as I have been taking courses and doing everything else. Although this independent research still needs a lot of work, it's in a good place, I have already improved many of the papers I have read and have multiple ideas to explore. In an ideal world I would win the lottery tomorrow and I could just support myself while I carry on with this research.

I am not sure how to approach my supervisor and tell him I don't want to finish this research, nor how to suggest to do my research instead and get a publication out of that, or even if I should touch on this topic at all.

  • "(...) which could take more than two weeks" - you mean, like 2.5 weeks, or 137 weeks? If the former, that's really not a big price for a publication, especially given the effort you've put in it already. Some people spend several months, years even, to get results. – user68958 Jun 14 '18 at 21:19
  • 5
    Side tip: Don't undercut your credentials/accomplishments. Warmly upsell. i.e.,: Cut the "quite pompous" deprecation. – Daniel R. Collins Jun 14 '18 at 21:23
  • 1
    @corey979 It's a bug, so in the worst case scenario I have to re-code the entire code base, which could take up to a month, plus improvement of the methods which is extra time. And the publication is not particularily impressive either, it was mostly taking my supervisors work and adding a couple of details to make it slightly more gneral. – Makogan Jun 14 '18 at 21:37
  • 1
    I don't think a lot of this question is necessary to understand the question, maybe you could pare it down a bit. – Azor Ahai Jun 14 '18 at 22:34
  • 3
    And how does this signaling help you? – Daniel R. Collins Jun 15 '18 at 0:23
3

Talk to your supervisor when you know how long you will need to finish the publication.

You are currently in the phase of a project where you have to make very sure everything really works. This is usually the least motivating stage, but it is absolutely necessary. So your supervisor will (hopefully) push you to go through this phase and finish the publication, because finishing research projects is something you have to learn as a budding researcher.

Of course, not every project should be finished. If the costs (time invested) of a project outweigh the benefits (results and publication strength), your supervisor may want to abandon the project. It is important to find out how long you still need, and only then (if still applicable - maybe it's a two-week bug fix that doesn't feel like drinking bleach) have a discussion with your supervisor. You can bring up your independent research during that discussion as well - either as an alternative or as a next project.

  • Thing is, i won't know until I have solved the bug depending on the severity I may need to rewrite the most critical sections of my code and that can take just as long as it originally took, which is months. – Makogan Jun 15 '18 at 8:56
  • I don't know your code of course, but for me the crucial step in determining how long I need to fix a bug is to identify the bug. That typically takes a lot shorter than the fixing itself. Maybe once you've identified the bug, you can say it will take a couple of weeks or a couple of months to fix? – Designerpot Jun 15 '18 at 9:06
  • The problem is, this was a bug that was discovered quite late. I had a set of tests and somehow none of them caught it, so I kept developing, doing my tests and verifying things, until I had to run the full simulation, and then the bug appeared (I could not run the full simulation until a lot of the code was made, hence why the tests were my only way of verifying that things worked). This means this bug has been there potentially since the beginning. Finding it is already scaring me, but my worst nightmare is finding it and realising how screwed everything is. – Makogan Jun 15 '18 at 9:32
0

Short answer: caffeine.

Long version: If you need to write code you will periodically be finding yourself in the frustrated and overwhelmed position of needing to return to square one to progress. For that matter pretty much all research gets like that regularly. The only way to avoid this frustrating state of affairs is to pay someone else to do it (warning, solution doesn't always work either).

Getting a very incremental project to start with, is pretty par for the course in grad school too. Nothing here to run screaming from.

It's also quite typical to hate a project after not seeing the progress you hoped for, particularly when it was punctuated with a big scary deadline. You can find a new adviser and/or start fresh with a project that has less baggage, but it will probably just happen again with something else sooner or later. If you like the area, you should stick with it; the frustration goes away as soon as you get some progress.

  • The frustration is related to the bug but the bug is not the cause, rather the frustration caused the bug, I was approaching the deadline, was both stressed and unmotivated by this point and needed to get things working fast so I ended up coding too quickly, which obviously generated horrible code. This hasn't happened with my other research that I have been working for for 6 months, despite that one also getting difficult at times. – Makogan Jun 15 '18 at 9:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.