Modifying the codes

[haghshenas]

We need to modify our codes so that they are consistent in how they get the input, generate output etc. This is useful for two reasons:

1. We eventually need to do this for the submission, so the earlier the better.
2. Since we are generating new simulations and we will do benchmarking and analysis, this consistency helps a lot and saves a lot of time (especially if we need to do the same on a new set of simulations)

My proposal is in the readme file of the repository, just check the following link:
https://github.com/haghshenas/tumorPhylogeny
I borrowed a lot from Simone's command line arguments style. So Simone should have the easiest job making the changes.

My focus was:

1. to have a single program for each of ILP, z3 and Open-WBO rather than two programs for simulated and real data
2. to have programs that are not needed to be changed every time we make small changes to the file names! That's why in my proposal the program should be fed with the noisy matrix (and the ground truth matrix, if available).

Please check it and let everyone know here if I have missed anything.
Otherwise, let's start to make changes in the codes.

[faridrashidi]

Great Ehsan. Two things:

1. Since our ground and noisy matrices we have now, I would suggest start col and row name index from 0. I mean mut0 mut1 ... and also cell0 cell1 ...
2. Do I need to add an argument for path z3?

[sciccolella]

The models for real data should provide a tree of the solution, or are we using something else based on the output?

[sciccolella]

The ILP code is updated.

[haghshenas]

1. Since our ground and noisy matrices we have now, I would suggest start col and row name index from 0. I mean mut0 mut1 ... and also cell0 cell1 ...

This is just naming. It should not matter in the code whether the cell name is cell0 or cell1.
I mean the code should treat all of these as strings!
Maybe we should emphasize in the readme that the output cell name and mutation name should be exactly the same as input. How about that?

2. Do I need to add an argument for path z3?

That's one way to go. But for now, lets keep it simple.
Have you tried the release binaries on z3 repository? I guess one of them should work on all linux machines.
Later we should add open-wbo and z3 as submodules ( I know how to do this, but it's a bit of hassle). This would also address the licensing issue, if there is any!