search

awpark / RCN_dimgen

0 stars 0 forks source link

Initial comments added #1

Open taddallas opened 7 years ago

taddallas commented 7 years ago

@awpark The manuscript looks really promising. I was careful to not really change too much of the text itself, and my comments are embedded in the Rmd file commented out html-style i.e., . Most of the comments are very "stream of consciousness". Given the number of comments you'll likely get on this draft, I can wait to provide more detailed comments. Also, if you have any particular aspect you'd like me to help with, let me know.

My edits are on the tad branch.

I did change the directory structure and the code to reproduce the figures, moving library calls to the beginning of the file, and making paths relative so that the directory is a bit more organized. This is just my OCD, and I apologize if I'm throwing a wrench into your repo structure.

The current csl file is for Biology Letters. Is this the target journal?

taddallas commented 7 years ago

Adding my comments on the most recent draft here (based on commit 94f0877).

It's looking solid. Toss some line numbers on it and tidy up the code for distribution and I'm ready to sign off on it. One small note about code. If you're planning on distributing through figshare or Dryad (I think), the repo structure has to be flat. I always forgot, and place things in folders (e.g., Data), and then kick myself later for it.

A couple of small comments:

"Closely related host species not only share common evolutionary history, and thus common parasites, at least until divergence in the more recent past [20,21], they also show higher similarity in their biology than distantly related hosts [22–24], including characteristics associated with immune responses, and behaviors that can modify potential for host switching [25–27]"

This sentence is a bit convoluted. Perhaps break into two?

Figure 3 and associated text in results.

Is it possible to compare the two distributions statistically? I'm thinking of something like the Kolmogorov–Smirnov test or maybe some likelihood ratio test magic?

subtending?

Had to Google that one. Is this the word you meant to use, and is there a more intuitive way to say this?

"specalist", "preivous", "specificty", and "exagerate" typos in discussion

Supplemental Table S1

Use xtable to produce pretty tables from R code. I know knitr has some functionality for this with the kable function, but I don't know if that gives you enough control over the table.