NEWS

mobr 3.0.2

Minor update

Decrease the amount of redundant documentation text by using inherit
Create functions get_samples and calc_comm_div_ci which together allow the user to compute confidence intervals for diversity metrics
Resurrect the old get_mob_stats function for making comparisons in diversity metrics between groups using confidence intervals and permutation tests.

Minor update

Major update

calc_comm_div now replaces get_mob_stats. The ancillary plotting function plot.mob_stats is now replaced with plot_comm_div. We hope to eventually add back in bootstrapping confidence intervals for the statistics but that is not currently supported. If you would like to know more about this design decision see the discussion here: https://github.com/MoBiodiv/mobr/issues/255.
plot_rarefaction is a bit more versatile with many new arguments that provide options for smoothing or averaging rarefaction curves when making comparisons within or between groups respectively.
The R package beta_C by Thore Engel has now been absorbed into the mobr package. Use function calc_beta_div or calc_comm_div with beta in the scales argument to compute coverage based beta diversity. Note the index in this case is S_C (i.e., richness S for a given level of coverage C)
a vignette was added to demonstrate how to make computations of beta diversity using mobr called beta_div_demo

remove sort on a data.frame from the function plot_abu which does not change how this function behaves but adheres to CRAN standards.

ref_level can be specified in get_mob_stats, plot_rarefaction and plot_sad which allows for the reference (i.e., control) group to be specified which all others are compared to. This is similar to the implementation in the function get_delta_stats

Gradient multi-scale mob method added to the function get_delta_stats as described in McGlinn et al. (2020
Added a new k-nearest neighbor algorithm for the spatial sample based rarefaction curve (sSBR)
ggplot integrated into plotting method for get_delta_stats -> plot.mob_out
Added automated tests
Improved documentation throughout the package

Function sphere_dist did not appear to be calculating great circle distances from latitude and longitude coordinates correctly using the Haversine formula. This was corrected with a new function this is correct according to comparisons with fields::rdist.earth and geosphere::distHaversine.

Removed the function overlap_effects which provided an overlapped image of the three components SAD, N, and agg across scale. This may be added back in the future

Refactored code surrounding the get_delta_stats function which computes the differences in richness across all possible scales.
Refactoring was primarily accomplished by integrating in the tidyverse tools which should make the code easier to understand and follow
Computational time has increased as the tidyverse tools are not necessarily the most computationally efficient solutions

provided core functionality of mob tools described in McGlinn et al. (2019) and Chase et al. (2018)