Private API

struct MergedTrajectory{F, C} <: Function

Callable struct merging a block-level trajectory base with sparse component-level overrides comps: f(t) evaluates base(t), normalises to a vector, applies each component override, and returns a scalar (dim == 1) or a vector.

base::F is a concrete type parameter, replacing the former base_fun::Function abstract capture. comps::C stores the Dict{Int,Function} override map.

Component index bounds are validated at construction, not at call time, so the call method stays allocation-free beyond what base and the component functions allocate.

Replaces the anonymous closure t -> begin … end (57 lines) produced inside CTModels.Init._build_block_with_components in builders.jl.

Fields

• base::F: Block-level trajectory function.
• comps::C: Component-level override map (Dict{Int,Function}).
• dim::Int: Total dimension.
• role::Symbol: Component role (:state or :control).

Examples

using CTModels.Init

base = t -> [0.0, 0.0]
comps = Dict{Int,Function}(2 => t -> sin(t))
f = MergedTrajectory(base, comps, 2, :state)
f(0.5)   # returns [0.0, sin(0.5)]

_build_block_with_components(
    ocp::CTModels.Models.AbstractModel,
    role::Symbol,
    block_data,
    comp_data::Dict{Int64, Any}
) -> Any

Build an initialisation function combining block-level and component-level data.

Arguments

• ocp::CTModels.Models.AbstractModel: The optimal control problem.
• role::Symbol: The component role (:state or :control).
• block_data: Block-level initialisation data.
• comp_data::Dict{Int,Any}: Component-level initialisation data indexed by component.

Returns

• Function: A combined initialisation function that merges block and component data.

Throws

• CTBase.Exceptions.IncorrectArgument: If dimensions are incompatible or component indices are out of bounds.

See also: CTModels.Init.MergedTrajectory, CTModels.Init.initial_state, CTModels.Init.initial_control

_build_component_function(data) -> Function

Build a component-level initialisation function from data.

Arguments

• data: The component data (time-dependent tuple or time-independent data).

Returns

• Function: A component initialisation function.

See also: CTModels.Init._build_component_function_without_time, CTModels.Init._build_component_function_with_time

_build_component_function_with_time(
    data,
    time::AbstractVector
) -> Function

Build a component function from data with an associated time grid.

Arguments

• data: The component data (function, scalar, or vector).
• time::AbstractVector: The time grid for interpolation.

Returns

• Function: A component initialisation function with time interpolation.

Throws

• CTBase.Exceptions.IncorrectArgument: If the data type is unsupported or time-grid mismatch occurs.

See also: CTModels.Components.ConstantInTime, CTBase.Interpolation.ctinterpolate

_build_component_function_without_time(data) -> Function

Build a component function from time-independent data (scalar, vector, or function).

Arguments

• data: The time-independent data (function, scalar, or vector).

Returns

• Function: A component initialisation function.

Throws

• CTBase.Exceptions.IncorrectArgument: If the data type is unsupported or vector length is invalid.

See also: CTModels.Components.ConstantInTime

_build_time_dependent_init(
    ocp::CTModels.Models.AbstractModel,
    role::Symbol,
    data,
    time::AbstractVector
) -> Union{Nothing, Function}

Build a time-dependent initialisation function from data and a time grid.

Arguments

• ocp::CTModels.Models.AbstractModel: The optimal control problem.
• role::Symbol: The component role (:state or :control).
• data: The data to interpolate (function, vector, or vector-of-vectors).
• time::AbstractVector: The time grid.

Returns

• Function: An interpolated initialisation function t -> value(t).

Throws

• CTBase.Exceptions.IncorrectArgument: If data type is unsupported or dimensions/time-grid mismatch occurs.

See also: CTBase.Interpolation.ctinterpolate, CTModels.Init.initial_state, CTModels.Init.initial_control

_format_init_data_for_grid(data) -> Any

Convert matrix data to vector-of-vectors format for time-grid interpolation.

Arguments

• data: The data to format (matrix or other format).

Returns

• Union{AbstractVector, Any}: The formatted data as vector-of-vectors, or the original data.

_format_time_grid(time_data) -> Any

Normalise time grid data to a vector format.

Arguments

• time_data: The time grid data (vector, array, or nothing).

Returns

• Union{Nothing, AbstractVector}: The normalised time grid as a vector, or nothing.

Throws

• Exceptions.IncorrectArgument: If the time grid type is invalid.

_initial_guess_from_namedtuple(
    ocp::CTModels.Models.AbstractModel,
    init_data::NamedTuple
) -> CTModels.Init.InitialGuess

Build an initial guess from a NamedTuple.

Arguments

• ocp::Models.AbstractModel: The optimal control problem.
• init_data::NamedTuple: The initial guess data with named fields.

Returns

• InitialGuess: An initial guess constructed from the NamedTuple.

Throws

• Exceptions.IncorrectArgument: If keys are invalid, dimensions mismatch, or specifications are duplicated.

_initial_guess_from_preinit(
    ocp::CTModels.Models.AbstractModel,
    pre::CTModels.Init.PreInitialGuess
) -> CTModels.Init.InitialGuess

Build an initial guess from a pre-initialisation object.

Arguments

• ocp::Models.AbstractModel: The optimal control problem.
• pre::PreInitialGuess: The pre-initialisation container.

Returns

• InitialGuess: An initial guess constructed from the pre-initialisation data.

_initial_guess_from_solution(
    ocp::CTModels.Models.AbstractModel,
    sol::CTModels.Solutions.AbstractSolution
) -> Union{CTModels.Init.InitialGuess{X, U, V} where {X<:Function, U<:Function, V<:Real}, CTModels.Init.InitialGuess{X, U, V} where {X<:Function, U<:Function, V<:(AbstractVector{<:Real})}}

Build an initial guess from a previous solution (warm start).

Arguments

• ocp::Models.AbstractModel: The optimal control problem.
• sol::Solutions.AbstractSolution: The previous solution.

Returns

• InitialGuess: An initial guess constructed from the solution.

Throws

• Exceptions.IncorrectArgument: If dimensions do not match between the problem and solution.

_validate_initial_guess(
    ocp::CTModels.Models.AbstractModel,
    init::CTModels.Init.InitialGuess
) -> CTModels.Init.InitialGuess

Internal validation of an InitialGuess.

Arguments

• ocp::Models.AbstractModel: The optimal control problem.
• init::InitialGuess: The initial guess to validate.

Returns

• InitialGuess: The validated initial guess.

Throws

• Exceptions.IncorrectArgument: If dimensions or types are incompatible.