This repository was archived by the owner on Sep 4, 2024. It is now read-only.
Rename "objective" to "trajectory" #72
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
goerz
force-pushed
the
rename-objectives
branch
from
6727f69 to
03e9a5d
Compare
goerz
force-pushed
the
rename-objectives
branch
from
03e9a5d to
cae4cb7
Compare
This was referenced Jan 22, 2024
goerz
added a commit
to JuliaQuantumControl/QuantumControl.jl
that referenced
this pull request
Jan 22, 2024
Renames `Objective` to `Trajectory`, `propagate_objective` to `propagate_trajectry`, the `objectives` keyword argument in `optimize` to `trajectories`, etc. This is a large, breaking change. The term "objective" was originally taken from the [`krotov` Python package](https://github.com/qucontrol/krotov), but has been a considerable source of misunderstanding. People (rightfully) get confused about "objectives" which are separate from the optimization functional. Especially when talking about a perfect entangler optimization where the objectives/trajectories have no target state, I always have to explain that 'objectives' is a misnomer, and that these are just the states whose dynamics enter the functional. The term "trajectory" is more accurate, as it consists of the initial state and the dynamical generator. Any (optional) additional properties (`target_state`, `weight`, …) are best understood as extra information specific to that trajectory that the functional may use. As part of this breaking change, there have been considerable revisions of related functionality. Trajectories and ControlProblems can now also be instantiated with positional arguments. Trajectories can now take propagation keyword arguments with a `prop_` prefix. These are taken into account in `propagate_trajectory` and the new `init_prop_trajectory`. These should be used in Krotov/GRAPE and finally enable customizing the propagation in an optimal control context (like a `prop_callback` that takes out population from the highest level, in lieu of a non-Hermitian Hamiltonian.) See also JuliaQuantumControl/QuantumControlBase.jl#72
goerz
added a commit
to JuliaQuantumControl/QuantumControl.jl
that referenced
this pull request
Jan 22, 2024
Renames `Objective` to `Trajectory`, `propagate_objective` to `propagate_trajectry`, the `objectives` keyword argument in `optimize` to `trajectories`, etc. This is a large, breaking change. The term "objective" was originally taken from the [`krotov` Python package](https://github.com/qucontrol/krotov), but has been a considerable source of misunderstanding. People (rightfully) get confused about "objectives" which are separate from the optimization functional. Especially when talking about a perfect entangler optimization where the objectives/trajectories have no target state, I always have to explain that 'objectives' is a misnomer, and that these are just the states whose dynamics enter the functional. The term "trajectory" is more accurate, as it consists of the initial state and the dynamical generator. Any (optional) additional properties (`target_state`, `weight`, …) are best understood as extra information specific to that trajectory that the functional may use. As part of this breaking change, there have been considerable revisions of related functionality. Trajectories and ControlProblems can now also be instantiated with positional arguments. Trajectories can now take propagation keyword arguments with a `prop_` prefix. These are taken into account in `propagate_trajectory` and the new `init_prop_trajectory`. These should be used in Krotov/GRAPE and finally enable customizing the propagation in an optimal control context (like a `prop_callback` that takes out population from the highest level, in lieu of a non-Hermitian Hamiltonian.) See also JuliaQuantumControl/QuantumControlBase.jl#72
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Renames
ObjectivetoTrajectory,propagate_objectivetopropagate_trajectry, theobjectiveskeyword argument inoptimizetotrajectories, etc.This is a large, breaking change.
The term "objective" was originally taken from the
krotovPython package, but has been a considerable source of misunderstanding. People (rightfully) get confused about "objectives" which are separate from the optimization functional. Especially when talking about a perfect entangler optimization where the objectives/trajectories have no target state, I always have to explain that 'objectives' is a misnomer, and that these are just the states whose dynamics enter the functional. The term "trajectory" is more accurate, as it consists of the initial state and the dynamical generator. Any (optional) additional properties (target_state,weight, …) are best understood as extra information specific to that trajectory that the functional may use.As part of this breaking change, there have been considerable revisions of related functionality. Trajectories and ControlProblems can now also be instantiated with positional arguments. Trajectories can now take propagation keyword arguments with a
prop_prefix. These are taken into account inpropagate_trajectoryand the newinit_prop_trajectory. These should be used in Krotov/GRAPE and finally enable customizing the propagation in an optimal control context (like aprop_callbackthat takes out population from the highest level, in lieu of a non-Hermitian Hamiltonian.)