16

During literature review for my research proposal, I have come across a professor who has published several papers in my field, some closely related to the topic of my own research. In a series of papers he uses a specific model, expanding on and extending it in each paper. In my work, I want to use the same kind of model (evolutionary) and I would love to see the source code, both to better understand and replicate the results and because it would tremendously help me in getting started. Unfortunately, I could not find it anywhere and it is not published with any of the papers.

I am wondering if and how I should ask for the code and have three related questions:

  1. Is it common to just email and ask for the source code or would that be considered somehow inappropriate? I might well end up applying to the university that professor is teaching at, so I don't want to leave a bad impression.

  2. Who is the appropriate person to contact? All of the papers on the model have been published by by the professor and the same two co-authors, some with additional co-authors. Should I email the professor himself, one of the co-authors, or all of them?

  3. How much detail about my own research should I include in my email? Just what's sufficient to justify the request or more, possibly a draft of the proposal?

  • 4
    @Solar Mike : He did not mention that he would "give nothing", either. Moreover, whether or not "something" should be "given" depends, at least I think, solely upon the other person. If they want something in return, then that must be honored. But if they are willing to give it for gratis, that too is their right. – The_Sympathizer Jan 18 at 1:38
  • 34
    @SolarMike It's sad to see that that is the mindset of a researcher these days. I mean, the code is part the paper, so what OP is asking for is something that should have actually been published with it. Fortunately, there has been some efforts on that front – user4052054 Jan 18 at 10:34
  • 10
    @SolarMike In my opinion, a paper based on code that does not contain such code (model, algorithm or whatever else) its incomplete. Is a psycology experiment paper valid without the data? Is a global temperature paper valid without the data? I hope not (even if there are many). I think a model/algorithm paper should not be valid without the code to reproduce the paper results. Research is (and should be open). – Ander Biguri Jan 18 at 11:38
  • 4
    @SolarMike At a bare minimum, he'd be giving the author of the original paper a citation. It's not much, but it's not nothing, either. – nick012000 Jan 18 at 12:05
  • 18
    @SolarMike Yes! NASA has tons of open-source code available at code.nasa.gov. Check out the projects called Computational Fluid Dynamics (CFD) Utility Software Library ("applicable to all flight regimes-subsonic, transonic, supersonic, and hypersonic"), TLNS3D ("efficient [...] for solving transonic viscous flows"), CFL3D, and Geometry Manipulation Protocol. – Vectornaut Jan 18 at 14:15
27

Whether common or not, it is acceptable to ask. Ask the main professor, who may pass on the request to a co-author.

I'd suggest that you only give a small amount of detail in the initial request but offer to say more if you like. But don't flood the professor in the initial request. Mention that you have a draft, but don't send it. And you can certainly mention that you are interested in applying to the university.

What happens next depends on the nature of the reply. Perhaps they will want to help you, but other outcomes are possible.

  • 3
    This is perfect; direct, honest and non-evasive. – AppliedAcademic Jan 18 at 12:11
  • 3
    My only amendment to this would be that it might be better to ask the author who is listed as a corresponding author on the paper - they can pass on the request as appropriate. – Ben Bolker Jan 18 at 21:24
9

Is it common to just email and ask for the source code or would that be considered somehow inappropriate?

It's not improper to ask. Consider some core values of the scientific world:

  • Reproducibility: for research to be truly valid, others should be able to reproduce it. This is why publishing source code is encouraged.

  • Standing on the shoulders of giants: science is all about building on each others' achievements. My discovery may help you and your discovery may help me. It's not a zero-sum game.

Who is the appropriate person to contact? All of the papers on the model have been published by by the professor and the same two co-authors, some with additional co-authors. Should I email the professor himself, one of the co-authors, or all of them?

Now, what is the best way to ask? You can email directly, but it would be nice if you can find a mutual acquaintance that can introduce you.

Is there anyone in your department (maybe your supervisor) that has met this professor before? One of the reasons scientists go to conferences is to build these networks. Or maybe someone in your department has co-authored something with the professor or his co-authors; go through the professor's publication list and look for familiar names.

Ask your supervisor if they know anything about the professor that's relevant. It's part of the supervisor's role to know the field and use it to advise you.

In the end, who you contact is a bit opportunistic. If you have a good connection to one particular person, start there.

How much detail about my own research should I include in my email? Just what's sufficient to justify the request or more, possibly a draft of the proposal?

I would keep it reasonably short. Enough to justify the request, certainly. But in general you don't share draft proposals until you've gained whatever grant you're applying for.

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.