Version 1.70.11

Version 1.70.10

Added matrix-free calculations for eigenvalue related calculations, for the eigenvector and the adjacency matrix related summary statistics. These are no longer restricted by tree size (although for very large trees, calculations might still take considerable time, scaling is much better now).
Added matrix-free calculations for the min_max laplacian function
Note that the RSpectra related statistics are NOT matrix free possible: these require all distances between all tips and all nodes, which can not be done matrix-free (so these are restricted to trees < ~20k nodes, and relatively slow).
Removed the large trees function as now we can calculate most statistics anyway for these large trees.

Added function to calculate statistics for large trees, which skips several distance matrix based statistics that can cause extreme memory usage

Removed all references and vignettes making use of the package abcrf, as this is longer available on CRAN.
Improved compliance to noSuggests CRAN policy.

Fixed numerical issue in calculating mean pair distance for very large trees
Now use armadillo to calculate Eigen values.
Added the function ‘sshape’, which was previously called the blum statistic (the blum function remains available as well). Thanks Sophie Kersting for pointing this out!
Polished the manual.

Added support for unrooted trees
Added check if tree is unrooted, and aborting if statistic does not support this
Updated reference table in the README file, showing which statistics are available for unrooted trees, and which statistics (appear) to be affected by root position
Using treebalance functions when treestats functions do not suffice (e.g. for rquartet for polytomous trees)
Updated Wiener test, to include distance matrix based calculation

Updated references.

Reduced dependencies, Matrix and RSpectra are no longer required (but availability will improve speed!)
calc_all_stats, calc_topology_stats and calc_brts_stats no longer return a named list, but return a named vector, for ease of rapid calculation across many trees

Removed dependencies to nodeSub by integrating functions to generate fully balanced and unbalanced into treestats.
Renamed ‘calc_balance_stats’ to ‘calc_topology_stats’, and included only statistics that take the topology (without the branch lengths) into account.

Renamed eigenvector to eigen_centrality (in line with the analogous function in igraph)

Add several new statistics:

Added checks to each statistic verifying the statistic requires an ultrametric or binary tree - this should avoid some rare instances where memory access violations would pop up when providing a non-binary tree to a statistic assuming a binary input phylogeny. Many thanks to Fien Strijthaegen for pointing this out.

Squashed a bug in imbalance_steps that would incorrectly calculate for smaller trees
Added bioRxiv link to DESCRIPTION
Added CITATION file

Corrected wording of variation in branch length statistics to correctly reflect variance, instead of variation