Review for alejandro_weinstein - *A Tale of Four Libraries*

[davclark]

Reviewer: Dav Clark Department/Center/Division: D-Lab Institution/University/Company: UC Berkeley Field of interest / expertise: Computational Social Science / Neuroscience Country: USA

Article reviewed: A Tale of Four Libraries

GENERAL EVALUATION

Please rate the paper using the following criteria (please use the abbreviation to the right of the description)::

below doesn't meet standards for academic publication meets meets or exceeds the standards for academic publication n/a not applicable

• Quality of the approach: meets
• Quality of the writing: meets
• Quality of the figures/tables: meets (though I find the standards for networks to be low!)

  SPECIFIC EVALUATION

For the following questions, please respond with 'yes' or 'no'. If you answer 'no', please provide a brief, one- to two-sentence explanation.

• Is the code made publicly available and does the article sufficiently describe how to access it?

  No - the github repo (https://github.com/aweinstein/a_tale) contains only a README. I suspect this is an oversight.

• Does the article present the problem in an appropriate context? Specifically, does it:

  • explain why the problem is important,

  Somewhat implicit - see below for questioning that "numpy arrays" in particular is what's important. The motivation for the science could be stronger.

  • describe in which situations it arises,

  Implicitly

  • outline relevant previous work,

  Theoretical, but not other approaches to similar theoretical concerns.

  • provide background information for non-experts

  Yes, but perhaps a bit technical (see below)

• Is the content of the paper accessible to a computational scientist with no specific knowledge in the given field?

  Yes, though the description of RL is somewhat technical for a broad audience. A formulation in which there is a decision rule and an update rule is perhaps less general, but likely easier to grok.

  Nice examples for similarity.

• Does the paper describe a well-formulated scientific or technical achievement?

  Yes

• Are the technical and scientific decisions well-motivated and clearly explained?

  Yes, but could be better motivated.

• Are the code examples (if any) sound, clear, and well-written?

  There are no code examples or code (I think the authors forgot to populate their companion github repo)

• Is the paper factually correct?

  As far as I can tell.

• Is the language and grammar of sufficient quality?

  Yes

• Are the conclusions justified?

  Yes, but they are a bit odd. I imagine few individuals making a choice about whether to use a library or not based on whether numpy arrays are used. Indeed, as long as a library supports the more generic python buffer interface, I suspect you'd have few problems compared to numpy arrays (e.g., with rpy2.robjects classes, PIL, etc.). Perhaps the strength comes from sparse numpy arrays?

• Is prior work properly and fully cited?

  Yes. It might be nice to see a comparison to alternate approaches, though.

• Should any part of the article be shortened or expanded? Please explain.

  Yes, at least some code snippets should be provided.

• In your view, is the paper fit for publication in the conference proceedings? Please suggest specific improvements and indicate whether you think the article needs a significant rewrite (rather than a minor revision).

  Yes

[aweinstein]

I populated the github repo: https://github.com/aweinstein/a_tale.

[davclark]

Cool - thanks!

If you had time to re-work "what's really important" that would be great, but as I said above, it's already good to publish in my book [pun intended].

[ahmadia]

This paper has been accepted, thanks.