Skip to content

Latest commit

 

History

History
43 lines (39 loc) · 3.46 KB

TODO.md

File metadata and controls

43 lines (39 loc) · 3.46 KB

TODO for dispersalDiversity

  • Continue on diversityMultilocus function returns match diversity function returns, also redo diversitySingleLocus returns
  • Do we need method = "permute"?
  • Update figures with existing code
  • continue generalising allele diversity tests
  • Clean up pairwise diversity tests, return values and print and plot methods
  • Make sites/groups/species terminology consistent
  • Consider adding other methods to createAlleleTables using the as.genalex functionality available to make conversions
  • copy chunks of diversity docs to diversityMultilocus and diversitySingleLocus
  • Note README includes several examples and more documentation on parameters and functions
  • Get data permission from VLS
  • Tests

Completed

  • Added a plot.divtable method which produces an annotated membershipPlot
  • membershipPlot annotation arguments l1 and l2 are placed by an annotation function which can be redefined by the user
  • Weighted means and sample variances of q.gg for each estimator and of q.gh, all as defined in Scofield et al. 2011 Oecologia, are now included in the results returned by diversity for divtable objects
  • For all diversity tests, the comparison of the log-likelihood value against an analytic Χ2 distribution is removed as this was not an appropriate test
  • Diversity tests now produce an error if any group has just one member, and as.divtable produces a warning for the same condition
  • membershipPlot writes the plot into a PDF file on option (the EPS option has been removed)
  • The pooled PMI values returned by diversity are documented
  • The divtable and allele_divtables classes have separate documentation
  • Export and document the nielsenTransform function
  • Production of pie plots and the method argument are removed from membershipPlot
  • The Gower distance matrix diagonal generation is much, much faster
  • Removed accum.method and distance.file arguments from gammaAccum functions. The proximity method was never well thought through; a proximity-based method could be valuable but not in the way it was implemented here.
  • For alpha and gamma diversity tests, all return class 'diversity_test' object which is handled by common print.diversity_test and plot.diversity_test methods
  • New as.divtable generic, with methods as.divtable.table, as.divtable.xtabs, as.divtable.matrix and as.divtable.data.frame, the latter of which first converts to matrix, then to divtable
  • Settled on new class divtable, which is shared with table
  • For allelic data, new class allele_divtables, which is shared with list and is a list of divtable objects for each locus
  • Quantiles checked are now richer and symmetric
  • Renamed allele.createTableList to S3 generic and method createAlleleTables and createAlleleTables.genalex. Added the synonym generic as.allele_divtables.
  • Stabilised function interfaces using S3 classes.
  • An example of random distance matrix creation is added to the README.
  • Reworked return value from diversity to return separate lists for q, q.nielsen and r
  • Changes function name of pmiDiversity to diversity
  • For data, added 2002 and 2004 granary assignments for Q. lobata
  • Worked out how to deal with library(RColorBrewer) use in membershipPlot. If RColorBrewer is available then it is used, with the new option fill.palette = "Dark2" selecting the palette. If it is not available, rainbow is used.