45 Matching Annotations
  1. Aug 2023
    1. Renee Grambihler

      Seems like good opportunity to network with Biosphere 2 & build DS capacity there. However, already pretty advanced practicioner.

    2. Rachel Gildersleeve

      Cut. No R

      This is Terrace's co-worker. We should follow up personally with resources to get started in R so they can take this course next year.

  2. Aug 2022
    1. However, it is also common to only include a subset of principal component scores when building regression models

      This is probably rarely a good idea. If ecologically relevant variables are not the ones that contribute to co-variation, they will be lost. In fact, principal component regression is rarely a good idea, especially since there are many supervised multivariate analysis techniques to deal with multicollinearity in regression like problems (e.g. RDA, CCA, PLSR). For more detailed discussion of why PCA regression is probably almost always the wrong choice for ecological data, see Scott & Crone 2021

      Scott, Eric R., and Elizabeth E. Crone. “Using the Right Tool for the Job: The Difference between Unsupervised and Supervised Analyses of Multivariate Ecological Data.” Oecologia 196 (February 12, 2021): 13–25. https://doi.org/10.1007/s00442-020-04848-w.

  3. Jul 2022
  4. Apr 2022
    1. Ecologists who write code often use the R programming language, and the rOpenSci community has a well-established software peer review process that involves both rOpenSci’s staff software engineers and the broader R user community. Their software review GitHub repository provides instructions for submitting an R package for review as well as guidelines for code reviewers. rOpenSci’s efforts have resulted in many well-used R packages for ecology research including rfishbase [21] and taxize [22].

      rOpenSci review is mentioned earlier in the Peer-Review section. I suggest moving this up and merging

  5. Mar 2022
    1. The standard GitHub licensing options are best suited for software. If your code is intended only for your specific analysis, consider a Creative Commons License. The Choose a License website can offer further guidance. If you wish to allow anyone to re-use your code, consider a CC0 1.0 public domain dedication. If you wish to receive attribution for any reuse of your code, consider a CC BY 4.0 license, which requires attribution upon reuse. If you have build an app, tool, package, or other product that you would like others to use and would like attribution for any reuse of your code, consider the GNU General Public License v3. This license also prohibits the re-user from making their re-used version private. If you do not wish to receive attribution and are open to private use, consider the MIT license.

      I think probably less detail is needed here. Distill down to most important points