baharev/sdopt-tearing

Exact and heuristic methods for tearing

sdopt-tearing's Issues

Any backend for Julia in perspective yet ?

Hello Ali,

Thanks first for providing us freely this codebase on the very specific subject of tearing.

I have seen there is some backend for AMPL, and Python.

I would like to know if it has been envisaged already to target the Julia language?
It's taking more and more place in the mathematical landscape.

FYI, the networkx package has been ported to lightgraphs.jl.

Cordially
Olivier

Interacting with the ordering algorithms

@baharev
I want to use sdopt-tearing in my Master's project but I am having some difficulty interacting with your code. Currently, I manually create the incidence matrix and pass it to the methods in heap_md.py to get the orderings using either the Hessenberg or Dulmage-Mendelsohn algorithms.

The problem that I have is ordering variables calling a function eg.

a = f(x, y, z)

My initial plan was to add the equation to the incidence matrix having the variable a, x, y, z but I run into situations where this equation is used (for example) to calculate the variable x.

I have read your documentation and see that there are methods that cater to for "bad" eliminations

determine which variables can be explicitly and safely eliminated from which equations

Is it possible for me to use this property to flag variables as bad eliminations? My plan is to create a new variable f and dummy equations eg.
f = x*y*z (1)
a = (x + y + z)*f (2)
and flag variable f as a "bad" elimination in equation 2. I create equation 1 to have the Degrees of Freedom of the system 0.

As a side note, these algorithms are extraordinary
Thanks in advance
Edgar

Add support for NumPy arrays

The following error is encountered when running

• rpc_api.fine_dulmage_mendelsohn
• heap_md.hessenberg
• bordered.to_bordered_form

ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

The errors originate from the following lines:

• dm_decomp.py line 148
• order_util.py line 104 + 105

CITATION file

I am currently part of a group of people writing a paper on testing for research software, as your code uses property based test (hypothesis) I would like to reference it. Have you thought about adding a citation file?

Details here: http://www.software.ac.uk/blog/2013-09-02-encouraging-citation-software-introducing-citation-files

Example here: https://github.com/dib-lab/khmer/blob/master/CITATION

Thanks!