Shuffle instance function

[CharJon]

I think we should add a testcase, that checks if writeProblem produces two different files.

[CharJon]

Also, I see a downside of this approach vs the model being shuffled before hand. If the shuffle is done by the parameters, calling "resetParams" on the model kind of removes the shuffle, which might be unexpected.

[codecov-io]

Codecov Report

Merging #81 (5750da1) into main (153bc70) will not change coverage. The diff coverage is 100.00%.

@@            Coverage Diff            @@
##              main       #81   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files           40        41    +1     
  Lines          694       701    +7     
=========================================
+ Hits           694       701    +7     

Impacted Files	Coverage Δ
geco/mips/utilities/generic.py	100.00% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 153bc70...bfb8f80. Read the comment docs.

[CharJon]

I'm not sure, if we should keep the test that uses optimize to check if shuffling works.

• It heavily depends on the yang instances and therefore has hidden side effects: If the yang implementation changes and the problem becomes easy (1 node) the test fails, even if shuffling might work.
• We are testing if the solution process changes and not if shuffling works. So from a testing theory perspective we are testing the wrong thing.
• It is slow (for a test).

If we want to keep a test like this, we should have a persistent test file in e.g. data/test-data

[mmghannam]

I'm not sure, if we should keep the test that uses optimize to check if shuffling works.

• It heavily depends on the yang instances and therefore has hidden side effects: If the yang implementation changes and the problem becomes easy (1 node) the test fails, even if shuffling might work.
• We are testing if the solution process changes and not if shuffling works. So from a testing theory perspective we are testing the wrong thing.
• It is slow (for a test).

If we want to keep a test like this, we should have a persistent test file in e.g. data/test-data

I agree, I will remove it.