Updates for New NL Interface

[ccoffrin]

@odow, can you give me a hand with how to update the tests of the form?

    func = constraint.func
    set = constraint.set
    @test isapprox(set.value, 0.64266 / get_base_flow(data); atol = 1e-4)
    @test isa(set, MOI.EqualTo{Float64})
    @test length(func.terms) == 1

[codecov]

Codecov Report

Attention: 6 lines in your changes are missing coverage. Please review.

Comparison is base (eda7791) 79.35% compared to head (73f285b) 79.35%.
Report is 1 commits behind head on master.

Files	Patch %	Lines
src/form/wp.jl	72.72%	3 Missing ⚠️
src/core/objective.jl	33.33%	2 Missing ⚠️
src/core/constraint_transient.jl	75.00%	1 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##           master     #256   +/-   ##
=======================================
  Coverage   79.35%   79.35%           
=======================================
  Files          39       39           
  Lines        5579     5579           
=======================================
  Hits         4427     4427           
  Misses       1152     1152           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[odow]

I'd need to take a deeper look. The tests fail because it looks like it's trying to use HiGHS for a nonlinear model???

GasModels.jl/test/debug.jl

Lines 27 to 29 in 5444b86

	@test length(func.terms) == 1
	var_ref = var(gm, :f_pipe)[128]
	@test isapprox(func.terms[var_ref], -1.0; atol = 1e-4)

This just looks like it is testing that func is -1.0 * var_ref?

I don't see why this would break with the changes.

[odow]

Just an FYI that I'm only online partially tomorrow, and then I'm away Fri/Sat/Sun NZ time so it may be next week before I get a proper look.

[ccoffrin]

No problem on the delay. From what I can tell the primary issue is this test @test isa(set, MOI.EqualTo{Float64}) probably the set type needs to change due to the new NL data structure.

[odow]

This was what I needed to get tests passing: lanl-ansi/Juniper.jl#263

diff --git a/test/runtests.jl b/test/runtests.jl
index f92f38f..c10df8b 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -34,6 +34,8 @@ juniper_solver = JuMP.optimizer_with_attributes(
     "nl_solver" => ipopt_solver,
     "mip_solver" => highs_solver,
     "log_levels" => [],
+    # TODO(odow): bug in Juniper
+    "feasibility_pump" => false,
 )
 
 

[odow]

This is the patch for the documentation

diff --git a/docs/Project.toml b/docs/Project.toml
index dfa65cd..3a52a5d 100644
--- a/docs/Project.toml
+++ b/docs/Project.toml
@@ -1,2 +1,5 @@
 [deps]
 Documenter = "e30172f5-a6a5-5a46-863b-614d45cd2de4"
+
+[compat]
+Documenter = "0.27"

[odow]

Doc fix is #257

Juniper just needs a new tag: lanl-ansi/Juniper.jl#266

[odow]

Suggested change

	Juniper = ">= 0.4"
	Juniper = "0.9.2"

Once lanl-ansi/Juniper.jl#266 is tagged

[odow]

@ccoffrin try rebasing this on master to get the doc fix, and bump the compat for Juniper