resolve conflicts

docs/Project.toml

@@ -1,7 +1,6 @@
 [deps]
 CDDLib = "3391f64e-dcde-5f30-b752-e11513730f60"
 Colors = "5ae59095-9a9b-59fe-a467-6f913c188581"
-Dionysos = "d92c97cf-b87d-42c1-a9c0-25df00b4d958"
 Documenter = "e30172f5-a6a5-5a46-863b-614d45cd2de4"
 FillArrays = "1a297f60-69ca-5386-bcde-b61e274b549b"
 GLPK = "60bf3e95-4087-53dc-ae20-288a0d20c6a6"

docs/make.jl

@@ -1,24 +1,40 @@
 using Dionysos
 using Documenter, Literate
 
-const EXAMPLES_DIR = joinpath(@__DIR__, "src", "examples")
+const EXAMPLES_SOLVERS_DIR = joinpath(@__DIR__, "src", "examples", "solvers")
+const EXAMPLES_UTILS_DIR = joinpath(@__DIR__, "src", "examples", "utils")
+
 const REFERENCE_DIR = joinpath(@__DIR__, "src", "reference")
 const OUTPUT_DIR = joinpath(@__DIR__, "src", "generated")
 
-const EXAMPLES = readdir(EXAMPLES_DIR)
+const EXAMPLES_SOLVERS = readdir(EXAMPLES_SOLVERS_DIR)
+const EXAMPLES_UTILS = readdir(EXAMPLES_UTILS_DIR)
 const REFERENCE = readdir(REFERENCE_DIR)
 
-for example in EXAMPLES
-    example_filepath = joinpath(EXAMPLES_DIR, example)
-    Literate.markdown(example_filepath, OUTPUT_DIR)
-    Literate.notebook(example_filepath, OUTPUT_DIR)
-    Literate.script(example_filepath, OUTPUT_DIR)
+function literate_actions(file, output_dir)
+    Literate.markdown(file, output_dir)
+    Literate.notebook(file, output_dir)
+    return Literate.script(file, output_dir)
+end
+
+for example in EXAMPLES_SOLVERS
+    literate_actions(joinpath(EXAMPLES_SOLVERS_DIR, example), OUTPUT_DIR)
+end
+for example in EXAMPLES_UTILS
+    literate_actions(joinpath(EXAMPLES_UTILS_DIR, example), OUTPUT_DIR)
 end
+literate_actions(joinpath(@__DIR__, "src", "examples", "Getting Started.jl"), OUTPUT_DIR)
 
 const _PAGES = [
     "Index" => "index.md",
     "Manual" => ["manual/abstraction-based-control.md", "manual/manual.md"],
-    "Examples" => map(EXAMPLES) do jl_file
+    "Getting Started" => "generated/Getting Started.md",
+    "Solvers" => map(EXAMPLES_SOLVERS) do jl_file
+        # Need `string` as Documenter fails if `name` is a `SubString{String}`.
+        name = string(split(jl_file, ".")[1])
+        return name => "generated/$name.md"
+    end,
+    "Utils" => map(EXAMPLES_UTILS) do jl_file
         # Need `string` as Documenter fails if `name` is a `SubString{String}`.
         name = string(split(jl_file, ".")[1])
         return name => "generated/$name.md"

docs/src/developers/setup.md

@@ -115,12 +115,13 @@ julia> using Revise
 
 If you don't plan to test the examples, comment out the Literate part in `docs/make.jl`:
 ```julila
- 11 #for example in EXAMPLES
- 12 #    example_filepath = joinpath(EXAMPLES_DIR, example)
- 13 #    Literate.markdown(example_filepath, OUTPUT_DIR)
- 14 #    Literate.notebook(example_filepath, OUTPUT_DIR)
- 15 #    Literate.script(example_filepath, OUTPUT_DIR)
- 16 #end
+ 20 # for example in EXAMPLES_SOLVERS
+ 21 #     literate_actions(joinpath(EXAMPLES_SOLVERS_DIR, example), OUTPUT_DIR)
+ 22 # end
+ 23 # for example in EXAMPLES_UTILS
+ 24 #     literate_actions(joinpath(EXAMPLES_UTILS_DIR, example), OUTPUT_DIR)
+ 25 # end
+ 26 # literate_actions(joinpath(@__DIR__, "src", "Getting Started.jl"), OUTPUT_DIR)
 ```
 This will speed up building the documentation quite a lot.
 

docs/src/examples/Getting Started.jl

@@ -23,7 +23,6 @@ const DI = Dionysos
 const UT = DI.Utils
 const DO = DI.Domain
 const ST = DI.System
-const CO = DI.Control
 const SY = DI.Symbolic
 
 # Additionally, we will short the submodules accondingly 

docs/src/examples/DC-DC converter.jl → docs/src/examples/solvers/DC-DC converter.jl

@@ -1,5 +1,5 @@
 using Test     #src
-# # Example: DC-DC converter
+# # Example: DC-DC converter solved by [Naive abstraction](https://github.com/dionysos-dev/Dionysos.jl/blob/master/docs/src/manual/manual.md#solvers).
 #
 #md # [![Binder](https://mybinder.org/badge_logo.svg)](@__BINDER_ROOT_URL__/generated/DC-DC converter.ipynb)
 #md # [![nbviewer](https://img.shields.io/badge/show-nbviewer-579ACA.svg)](@__NBVIEWER_ROOT_URL__/generated/DC-DC converter.ipynb)
@@ -19,24 +19,13 @@ using Test     #src
 # A_1 = \begin{bmatrix} -\frac{r_l}{x_l} &0 \\ 0 & -\frac{1}{x_c}\frac{1}{r_0+r_c}  \end{bmatrix}, A_2= \begin{bmatrix} -\frac{1}{x_l}\left(r_l+\frac{r_0r_c}{r_0+r_c}\right) & -\frac{1}{x_l}\frac{r_0}{r_0+r_c}  \\ \frac{1}{x_c}\frac{r_0}{r_0+r_c}   & -\frac{1}{x_c}\frac{1}{r_0+r_c}  \end{bmatrix}, b = \begin{bmatrix} \frac{v_s}{x_l}\\0\end{bmatrix}.
 # ```
 # The goal is to design a controller to keep the state of the system in a safety region around the reference desired value, using as input only the switching
-# signal.
-#
-#
-# In order to study the concrete system and its symbolic abstraction in a unified framework, we will solve the problem
-# for the sampled system with a sampling time $\tau$.
-#
-# The abstraction is based on a feedback refinment relation [4,V.2 Definition].
-# Basically, this is equivalent to an alternating simulation relationship with the additional constraint that the input of the
-# concrete and symbolic system preserving the relation must be identical.
-# This allows to easily determine the controller of the concrete system from the abstraction controller by simply adding a quantization step.
-#
-# For the construction of the relations in the abstraction, it is necessary to over-approximate attainable sets of
-# a particular cell. In this example, we consider the used of a growth bound function  [4, VIII.2, VIII.5] which is one of the possible methods to over-approximate
-# attainable sets of a particular cell based on the state reach by its center. Therefore, it is used
-# to compute the relations in the abstraction based on the feedback refinement relation.
+# signal. In order to study the concrete system and its symbolic abstraction in a unified framework, we will solve the problem
+# for the sampled system with a sampling time $\tau$. For the construction of the relations in the abstraction, it is necessary to over-approximate attainable sets of
+# a particular cell. In this example, we consider the use of a growth bound function  [4, VIII.2, VIII.5] which is one of the possible methods to over-approximate
+# attainable sets of a particular cell based on the state reach by its center.
 #
 
-# First, let us import [StaticArrays](https://github.com/JuliaArrays/StaticArrays.jl) and [Plots].
+# First, let us import [StaticArrays](https://github.com/JuliaArrays/StaticArrays.jl) and [Plots](https://github.com/JuliaPlots/Plots.jl).
 
 using StaticArrays, Plots
 
@@ -47,7 +36,6 @@ const UT = DI.Utils
 const DO = DI.Domain
 const ST = DI.System
 const SY = DI.Symbolic
-const CO = DI.Control
 const OP = DI.Optim
 const AB = OP.Abstraction
 
@@ -67,7 +55,7 @@ hu = SVector(1)
 input_grid = DO.GridFree(u0, hu)
 
 using JuMP
-optimizer = MOI.instantiate(AB.SCOTSAbstraction.Optimizer)
+optimizer = MOI.instantiate(AB.NaiveAbstraction.Optimizer)
 MOI.set(optimizer, MOI.RawOptimizerAttribute("concrete_problem"), concrete_problem)
 MOI.set(optimizer, MOI.RawOptimizerAttribute("state_grid"), state_grid)
 MOI.set(optimizer, MOI.RawOptimizerAttribute("input_grid"), input_grid)
@@ -82,12 +70,12 @@ concrete_controller = MOI.get(optimizer, MOI.RawOptimizerAttribute("concrete_con
 # as well as the true initial state `x0` which is contained in the initial state-space defined previously.
 nstep = 300
 x0 = SVector(1.2, 5.6)
-x_traj, u_traj =
-    CO.get_closed_loop_trajectory(concrete_system.f, concrete_controller, x0, nstep)
+control_trajectory =
+    ST.get_closed_loop_trajectory(concrete_system.f, concrete_controller, x0, nstep)
 
 fig = plot(; aspect_ratio = :equal);
 plot!(concrete_system.X);
-plot!(UT.DrawTrajectory(x_traj))
+plot!(control_trajectory)
 
 # ### References
 # 1. A. Girard, G. Pola and P. Tabuada, "Approximately Bisimilar Symbolic Models for Incrementally Stable Switched Systems," in IEEE Transactions on Automatic Control, vol. 55, no. 1, pp. 116-126, Jan. 2010.

docs/src/examples/Gol, Lazar & Belta (2013).jl → docs/src/examples/solvers/Gol, Lazar & Belta (2013).jl

@@ -1,5 +1,5 @@
 using Test     #src
-# # Example: Gol, Lazar and Belta (2013)
+# # Example: Gol, Lazar and Belta (2013) solved by [Bemporad Morari](https://github.com/dionysos-dev/Dionysos.jl/blob/master/docs/src/manual/manual.md#solvers).
 #
 #md # [![Binder](https://mybinder.org/badge_logo.svg)](@__BINDER_ROOT_URL__/generated/Gol%2C Lazar %26 Belta (2013).ipynb)
 #md # [![nbviewer](https://img.shields.io/badge/show-nbviewer-579ACA.svg)](@__NBVIEWER_ROOT_URL__/generated/Gol%2C Lazar %26 Belta (2013).ipynb)
@@ -30,7 +30,7 @@ import Ipopt
 using Dionysos
 const DI = Dionysos
 const UT = DI.Utils
-const CO = DI.Control
+const ST = DI.System
 const OP = DI.Optim
 
 # And the file defining the hybrid system for this problem
@@ -86,7 +86,7 @@ objective_value = MOI.get(optimizer, MOI.ObjectiveValue())
 @test objective_value ≈ 11.38 atol = 1e-2     #src
 
 # and recover the corresponding continuous trajectory
-xu = MOI.get(optimizer, CO.ContinuousTrajectoryAttribute());
+xu = MOI.get(optimizer, ST.ContinuousTrajectoryAttribute());
 
 # ## A little bit of data visualization now:
 

docs/src/examples/Hierarchical-abstraction.jl → docs/src/examples/solvers/Hierarchical-abstraction.jl

@@ -1,4 +1,4 @@
-# # Hierarchical-abstraction
+# # Example: Reachability problem solved by [Hierarchical abstraction](https://github.com/dionysos-dev/Dionysos.jl/blob/master/docs/src/manual/manual.md#solvers).
 #
 #md # [![Binder](https://mybinder.org/badge_logo.svg)](@__BINDER_ROOT_URL__/generated/Hierarchical-abstraction.ipynb)
 #md # [![nbviewer](https://img.shields.io/badge/show-nbviewer-579ACA.svg)](@__NBVIEWER_ROOT_URL__/generated/Hierarchical-abstraction.ipynb)
@@ -13,12 +13,11 @@ const UT = DI.Utils
 const DO = DI.Domain
 const ST = DI.System
 const SY = DI.Symbolic
-const CO = DI.Control
 const PR = DI.Problem
 const OP = DI.Optim
 const AB = OP.Abstraction
 
-include("../../../problems/simple_problem.jl")
+include(joinpath(dirname(dirname(pathof(Dionysos))), "problems", "simple_problem.jl"))
 
 ## specific functions
 function post_image(abstract_system, concrete_system, xpos, u)
@@ -126,11 +125,11 @@ AB.LazyAbstraction.set_optimizer_parameters!(
 )
 
 # Global optimizer parameters
-hx_global = [10.0, 10.0] #[15.0, 15.0]
+hx_global = [10.0, 10.0]
 u0 = SVector(0.0, 0.0)
 hu = SVector(0.5, 0.5)
 Ugrid = DO.GridFree(u0, hu)
-max_iter = 6 # 9
+max_iter = 6
 max_time = 1000
 
 optimizer = MOI.instantiate(AB.HierarchicalAbstraction.Optimizer)
@@ -160,9 +159,12 @@ println("Solved : ", optimizer.solved)
 
 # ## Simulation
 x0 = UT.get_center(concrete_problem.initial_set)
-x_traj, u_traj, cost_traj = AB.HierarchicalAbstraction.simulate_trajectory(optimizer, x0)
-cost = sum(cost_traj);
-println("Goal set reached: $(x_traj[end]∈concrete_problem.target_set)")
+cost_control_trajectory =
+    AB.HierarchicalAbstraction.get_closed_loop_trajectory(optimizer, x0)
+cost = sum(cost_control_trajectory.costs.seq);
+println(
+    "Goal set reached: $(ST.get_state(cost_control_trajectory, ST.length(cost_control_trajectory))∈concrete_problem.target_set)",
+)
 println("Cost:\t $(cost)")
 
 # ## Display the results
@@ -180,7 +182,7 @@ plot!(concrete_problem.initial_set; color = :green, opacity = 0.8);
 plot!(concrete_problem.target_set; dims = [1, 2], color = :red, opacity = 0.8);
 
 #We display the concrete trajectory
-plot!(UT.DrawTrajectory(x_traj); ms = 0.5)
+plot!(cost_control_trajectory; ms = 0.5)
 
 # # Display the lazy abstraction 
 fig = plot(; aspect_ratio = :equal);
@@ -191,4 +193,4 @@ plot!(
     heuristic = false,
     fine = true,
 )
-plot!(UT.DrawTrajectory(x_traj); ms = 0.5)
+plot!(cost_control_trajectory; ms = 0.5)

docs/src/examples/Lazy-Ellipsoids-Abstraction.jl → docs/src/examples/solvers/Lazy-Ellipsoids-Abstraction.jl

@@ -1,3 +1,6 @@
+# # Example: Reachability problem solved by [Lazy ellipsoid abstraction](https://github.com/dionysos-dev/Dionysos.jl/blob/master/docs/src/manual/manual.md#solvers).
+#
+
 using StaticArrays, LinearAlgebra, Random, IntervalArithmetic
 using MathematicalSystems, HybridSystems
 using JuMP, Mosek, MosekTools
@@ -11,12 +14,11 @@ const UT = DI.Utils
 const DO = DI.Domain
 const ST = DI.System
 const SY = DI.Symbolic
-const CO = DI.Control
 const PR = DI.Problem
 const OP = DI.Optim
 const AB = OP.Abstraction
 
-include("../../../problems/non_linear.jl")
+include(joinpath(dirname(dirname(pathof(Dionysos))), "problems", "non_linear.jl"))
 
 # # First example
 
@@ -73,7 +75,7 @@ reached(x) = x ∈ concrete_problem.target_set
 nstep = typeof(concrete_problem.time) == PR.Infinity ? 100 : concrete_problem.time; # max num of steps
 # We simulate the closed loop trajectory
 x0 = concrete_problem.initial_set.c
-x_traj, u_traj, cost_traj = CO.get_closed_loop_trajectory(
+cost_control_trajectory = ST.get_closed_loop_trajectory(
     concrete_system.f_eval,
     concrete_controller,
     cost_eval,
@@ -83,7 +85,7 @@ x_traj, u_traj, cost_traj = CO.get_closed_loop_trajectory(
     noise = true,
 )
 cost_bound = concrete_lyap_fun(x0)
-cost_true = sum(cost_traj);
+cost_true = sum(cost_control_trajectory.costs.seq);
 println("Goal set reached")
 println("Guaranteed cost:\t $(cost_bound)")
 println("True cost:\t\t $(cost_true)")
@@ -96,23 +98,24 @@ fig = plot(;
     ytickfontsize = 10,
     guidefontsize = 16,
     titlefontsize = 14,
+    label = false,
 );
 xlabel!("\$x_1\$");
 ylabel!("\$x_2\$");
 title!("Specifictions and domains");
 
 #Display the concrete domain
-plot!(concrete_system.X; color = :yellow, opacity = 0.5);
+plot!(concrete_system.X; color = :yellow, opacity = 0.5, label = false);
 for obs in concrete_system.obstacles
-    plot!(obs; color = :black)
+    plot!(obs; color = :black, label = false)
 end
 
 #Display the abstract domain
-plot!(abstract_system; arrowsB = false, cost = false);
+plot!(abstract_system; arrowsB = false, cost = false, label = false);
 
 #Display the concrete specifications
-plot!(concrete_problem.initial_set; color = :green);
-plot!(concrete_problem.target_set; color = :red)
+plot!(concrete_problem.initial_set; color = :green, label = false);
+plot!(concrete_problem.target_set; color = :red, label = false)
 
 # # Display the abstraction
 fig = plot(;
@@ -141,6 +144,6 @@ for obs in concrete_system.obstacles
     plot!(obs; color = :black)
 end
 plot!(abstract_system; arrowsB = false, cost = true);
-plot!(UT.DrawTrajectory(x_traj); color = :black)
+plot!(cost_control_trajectory; color = :black)
 
 @test cost_true <= cost_bound             #src