#20 - Add assignment as getter function of reset map

[mforets]

Closes #20.

[blegat]

guard could use assignment in his implementation

[codecov]

Codecov Report

Merging #22 into master will increase coverage by 0.11%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master      #22      +/-   ##
==========================================
+ Coverage   81.13%   81.25%   +0.11%     
==========================================
  Files           6        6              
  Lines         159      160       +1     
==========================================
+ Hits          129      130       +1     
  Misses         30       30

Impacted Files	Coverage Δ
src/HybridSystems.jl	71.42% <100%> (+2.19%)	⬆️

---

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 a2d1a9f...a4d11da. Read the comment docs.

[schillic]

I find this implementation misleading. In the standard terminology, the assignment is the part of the transition that is applied after taking the transition. In this package, resetmaps wraps both the (usual) assignment and the guard.

[blegat]

That's a good point, what do you suggest ?

[mforets]

I see what you mean; but the "problem" is that the maps are types, not functions. (we could otherwise let assignment return the apply function ... ).

Consider here page 9 (which refers to Henzinger's), it calls it "jump relation" Jump(e) as (guard, assignment) tuples for each edge.

[schillic]

Yes, I would change the name. "Jump relation" could also be misunderstood as the transition relation. resetmap would be in the terminology of this package.

[mforets]

Just that? To let resetmap(hs, t) = hs.resetmaps[symbol(hs, t)] ? And then we can do apply(resetmap(hs, t), X)

And how about assignment(hs, t) = x -> apply(hs.resetmaps[symbol(hs, t)], x...)?

[schillic]

Hm, thinking about this, I kind of disagree again. guard is a getter, while assignment is proposed to apply the assignment. How about apply_assignment instead?

[mforets]

assignment doesn't apply the assignment; it returns the function. So i understand it also as a getter. It is different from assignment(hs, t, args).

[schillic]

You wrote assignment(hs, t) = x -> apply(hs.resetmaps[symbol(hs, t)], x...) above.
Are you saying that apply does not apply? 🙀

EDIT: I see now that you mean the change in #25. Then I agree again.