[Doc] Relecture de la documentation

[ocots]

Cette issue a pour objectif de recenser les modifications à apporter au code suite à la relecture de la documentation.

[joseph-gergaud]

Basic exemple

Exporting and importing the solution

The codes don't work 👍 using JSON3

julia> export_ocp_solution(sol; filename="my_solution", format=:JSON)
ERROR: ExtensionError. Please make: julia> using JLD2, JSON3
Stacktrace:
[1] export_ocp_solution(args::OptimalControlSolution; kwargs::@kwargs{filename::String, format::Symbol})
...

[ocots]

Basic exemple

Exporting and importing the solution

The codes don't work 👍 using JSON3

julia> export_ocp_solution(sol; filename="my_solution", format=:JSON) ERROR: ExtensionError. Please make: julia> using JLD2, JSON3 Stacktrace: [1] export_ocp_solution(args::OptimalControlSolution; kwargs::@kwargs{filename::String, format::Symbol}) ...

@joseph-gergaud Have you updated the packages? The message

Please make: julia> using JLD2, JSON3

is weird since now JLD2 and JSON3 are split:

https://github.com/control-toolbox/CTModels.jl/blob/2e4b2caaa49b093407f533b8771101466ba62900/src/CTModels.jl#L133-L153

[jbcaillau]

@joseph-gergaud @ocots better if we contain this example in a single doc page, not to block the new release for sth that is rather a detail (and most probably used by very few)

[joseph-gergaud]

Define a problem

1. The user don’t use the following syntax
   :( $v ∈ R^$q, variable )
   :( $v ∈ R , variable )

So, why do we mention it ?

Constraints
Exemple
for me there is 4 linear boundary constraints and one nonlinear one-sided not two-sided) constraint

:
• two linear boundary constraints,
• one linear variable constraint,
• one linear state constraint,
• one (two-sided) nonlinear control constraint.

[ocots]

Define a problem

1. The user don’t use the following syntax
   :( $v ∈ R^$q, variable )
   :( $v ∈ R , variable )

So, why do we mention it ?

Constraints Exemple for me there is 4 linear boundary constraints and one nonlinear one-sided not two-sided) constraint

: • two linear boundary constraints, • one linear variable constraint, • one linear state constraint, • one (two-sided) nonlinear control constraint.

Il y a bien 2 boundary pour x(0) et x(tf), 1 linéaire sur la variable tf, 1 linéaire sur x(t) et 1 non linéaire sur u.

Par contre elle était bien "one-sided", j'ai modifié ça : #569.

[joseph-gergaud]

On a deux fois

time : t, ....

julia> @def damped_integrator begin
           tf ∈ R, variable
           t ∈ [0, tf], time
           x = (q, v) ∈ R², state
           u ∈ R, control
           q̇ = v(t)
           v̇ = u(t) - c(t)
           ẋ(t) == [q̇, v̇]
       end true;
variable: tf, dim: 1
time: t, initial time: 0, final time: var"tf##616"[1]
state: x, dim: 2
control: u, dim: 1
alias: q̇ = (x[2])(t)
alias: v̇ = (var"u##617"[1])(t) - c(t)
dynamics: ∂(x)(t) == [(x[2])(t), (var"u##617"[1])(t) - c(t)]
variable: tf, dim: 1
time: t, initial time: 0, final time: var"tf##632"[1]
state: x, dim: 2
control: u, dim: 1
alias: q̇ = (x[2])(t)
alias: v̇ = (var"u##635"[1])(t) - c(t)
dynamics: ∂(x)(t) == [(x[2])(t), (var"u##635"[1])(t) - c(t)]
ERROR: UnauthorizedCall: the objective must be set before building the model.

[ocots]

On a deux fois

time : t, ....
julia> @def damped_integrator begin
tf ∈ R, variable
t ∈ [0, tf], time
x = (q, v) ∈ R², state
u ∈ R, control
q̇ = v(t)
v̇ = u(t) - c(t)
ẋ(t) == [q̇, v̇]
end true;
variable: tf, dim: 1
time: t, initial time: 0, final time: var"tf##616"[1]
state: x, dim: 2
control: u, dim: 1
alias: q̇ = (x[2])(t)
alias: v̇ = (var"u##617"[1])(t) - c(t)
dynamics: ∂(x)(t) == [(x[2])(t), (var"u##617"[1])(t) - c(t)]
variable: tf, dim: 1
time: t, initial time: 0, final time: var"tf##632"[1]
state: x, dim: 2
control: u, dim: 1
alias: q̇ = (x[2])(t)
alias: v̇ = (var"u##635"[1])(t) - c(t)
dynamics: ∂(x)(t) == [(x[2])(t), (var"u##635"[1])(t) - c(t)]
ERROR: UnauthorizedCall: the objective must be set before building the model.

Celle-là, elle pour @jbcaillau je pense. CTParser ?

[jbcaillau]

@joseph-gergaud what is the issue? (apart from the fact that the cost is not defined.)

[ocots]

@joseph-gergaud what is the issue? (apart from the fact that the cost is not defined.)

The output is printed twice.