A lot of scientific research code is execute-once-throw-away code. Once the code obtains a result you can put into your research paper, it has served its purpo

A lot of scientific research code is execute-once-throw-away code. Once the code obtains a result you can put into your research paper, it has served its purpose and it'll never be used again. There is no need for any kind of abstraction, software design, architecture or development best practices. You're not going to be supporting this code or extending it in the future. If you're a software developer, the way researchers write code rubs you the wrong way. But you have to accept the reality. Applying software development principles to scientific code is just a waste of time.