NEWS.md
In mailund/admixture_graph: Admixture Graph Manipulation and Fitting

Version 1.0.2.9000

Fixed a bug in the MCMC code.
Introduced admixture graph library in the data folder in the form of matrices, and new functions graph_to_vector() and vector_to_graph() to interpret the matrix data.
Added rename_nodes() for renaming graph nodes in user defined or automised manner.
Added canonise_graph() and remove_duplicates() for detecting isomorphism among graphs with labeled leaves.
Added all_graphs() although the data folder already contains a big admixture graph library.
New parallel computing fitting function fit_graph_list() that takes one list and data as an input instead of a list and permutations.
Updated the vignette.

Updated code for laying out graphs with nodes that have more than two children.
Added options for fitting plots to make them in grayscale (Issue #11).
Updated functions for plotting the cost function and number of fits as functions of admixture proportions (Issue #12).
Changed specification of admixture proportions (Iussue #9) to avoid specifying redundant information. The changes are still backward compatible, so old analysis scripts should still work.

Fixed bug that caused crossing edges for admixture events in plotting (Issue #3).
Added bears to data/
Added syntactic suggar to graph construction
Fix plotting bug (Issue #4)
Wrote an interface to fitted data objects using the generic functions usually used for fitted models.
Added code for plotting fitted data
Added predict() function.
Added function for filtering data based on the leaves in a tree and for mapping between samples and populations (leaves).
New optimisation code based on solving linear systems for those variables that are linear.