More SpineOpt

[joaquimg]

• Add info in main readme
• Case simplified to run a single month (instead of 12)
• @datejada there are still some non-smooth install issues: (Python) Installation issue spine-tools/SpineOpt.jl#1142

[joaquimg]

There is something non-deterministic.
Every time I run I get a different Hash.
Is this expected @datejada ?

[joaquimg]

Row labels get messed up:

[datejada]

There is something non-deterministic. Every time I run I get a different Hash. Is this expected @datejada ?

@joaquimg I'm not aware that something is non-deterministic. @manuelma or @Tasqu, do you have any idea about this behaviour in SpineOpt? Basically, the order of the constraints in the .mps file seems to change for each run.

[joaquimg]

@odow, can you test running the SpineOpt script on your side to see if everything is properly documented?

[Tasqu]

@datejada I'm not surprised it's non-deterministic. We use Julia Dicts in SpineOpt quite a bit and their ordering is not consistent. Furthermore, I talked with Antti, and depending on how SpineInterface queries the databases it might also go through Python dictionaries which are not ordered either.

[joaquimg]

I am merging this, as it is a strict improvement over what we currently have.