gruph

Fits graphical models using kruskal's algorithm. At the moment it supports

• calculating raw edges for gaussian, discrete, and mixed cases.
• setting a threshold

Todo

• make secure with Principals and mechanisms https://www.securecoding.cert.org/confluence/pages/viewpage.action?pageId=637
• get working on parellel version.
• write documentation

assertions

• change ints to size_t
• make test for column assertion
• handle NA NaN and Inf
• test that discrete elements aren't unique!
• make tests for the above

functionality

• add model for degenerate gaussian
• parallel
• refacture threshold_init and threshold_initializer

testing

• compare vanilla kruskal to grapHD.

read up on

• Pretty code with iterators

• https://www.cs.helsinki.fi/u/tpkarkka/alglib/k06/

• BOOST accumulators

• BOOST for each

• BOOST graph