Commdyn Weekly Meeting

Participants: Matt, Chris, Lauren W., Syd, Peter, Lauren, H., Scott, Andrew, Corinna

Agenda and notes:

Codyn package discussion

• Lauren - finished variance ratio, starting turnover metrics
• Andrew - Taylor series https://github.com/laurenmh/codyn/issues/5
• Syd - timelag analysis
• TODO before first release
  • Documentation needs improvement for all methods
    • describe in documentation that replicate column should be unique.
  • Simpler metrics: general community stability
  • Better null model testing; incorporate that into their final forms
  • better Vignettes
  • define a class with print method?
  • rank clocks: Syd can tackle this
  • rank shift
  • TODO: Travis CI needs to be turned on by Lauren
    • See for example: https://github.com/DataONEorg/rdataone
• add tests for all new functions.
  • Scott: cumulative turnover plots would be nice to add, Syd has functional code for these turnover plots
  • 15 metrics on wikipedia http://en.wikipedia.org/wiki/Taylor%27s_law
  • Package not currently loading; problem with timelag_analysis.R not in functions
  • Two functions:
    • VR, which gives the option of two (or more?) null models for testing
    • TPL, allowing people to pass a formula to a modelling function (lme), providing a formula object, random effects (if any). Outputs model object, residuals, and a relationship between resid and dominance.
  • Add tests for
    • Variance Ratio [DONE?]
    • Taylor power law
    • Turnover metrics
    • Timelag metrics (Collins et al. 2000) http://temperate.lternet.edu/collins/biblio/disturbance-frequency-and-community-stability-native-tallgrass-prairie
    • Rank Shifts
  • Build completeness
    • Matt: Roxygen
    • Matt: NAMESPACE
• Next discussion
  • continue discussions in github issues
  • 3 weeks from now (Nov 14)
  • Target first release end of November

RecordR

- Need feedback form science team
- Will ask for feedback next week during Oct 31 - thanks.