[Numpy-discussion] The future of ndarray.diagonal()

[Konrad Hinsen]

On 04/01/15 21:28, Sturla Molden wrote:

> A scientific paper or thesis should be written so it is completely
> reproducible. That would include describing the computer, OS, Python
> version and NumPy version, as well as C or Fortran compiler.

I completely agree and we should all work towards this goal. But we 
aren't there yet. Most of the scientific community is just beginning to 
realize that there is a problem. Anyone writing scientific software for 
use in today's environment has to take this into account.

More importantly, there is not only the technical problem of 
reproducibility, but also the meta-level problem of human understanding. 
Scientific communication depends more and more on scripts as the only 
precise documentation of a computational method. Our programming 
languages are becoming a major form of scientific notation, alongside 
traditional mathematics. Humans don't read written text with version 
numbers in mind.

This is a vast problem which can't be solved merely by "fixing" software 
technology, but it's something to keep in mind nevertheless when writing 
software. For those interested in this aspect, I have written a much 
more detailed account in a recent paper:

	http://dx.doi.org/10.12688/f1000research.3978.2

Konrad.