Refactor SemOptimizer API #299

alyst · 2026-01-27T18:58:57Z

Extracted from #245 -- moving all engine-specific code (types) into extensions for modularity and leaner API, updating docstrings to make all optimizer help available for the user.

Requires #245 to be merged first.

src/optimizer/abstract.jl

src/optimizer/optim.jl

codecov · 2026-01-27T19:08:00Z

Codecov Report

❌ Patch coverage is 56.25000% with 14 lines in your changes missing coverage. Please review.
✅ Project coverage is 72.15%. Comparing base (8b0f880) to head (2b42351).
⚠️ Report is 58 commits behind head on devel.

Files with missing lines	Patch %	Lines
src/optimizer/abstract.jl	15.38%	11 Missing ⚠️
ext/SEMNLOptExt/NLopt.jl	86.66%	2 Missing ⚠️
src/optimizer/Empty.jl	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##            devel     #299      +/-   ##
==========================================
- Coverage   72.94%   72.15%   -0.80%     
==========================================
  Files          50       51       +1     
  Lines        2218     2216       -2     
==========================================
- Hits         1618     1599      -19     
- Misses        600      617      +17

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

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

This reverts commit 9729819.

This reverts commit 81a4bd9.

This reverts commit f0df653.

It is a simple and intuitive syntax and avoids declaring new types. Also allow specifying default constraint tolerance as `constraint_tol`.

use SemOptimizer(engine = :NLopt) instead of SemOptimizerNLopt() as this is a more universal scheme

It works around docstring errors during SemOptimizer{:E} parsing.

Instead of engine-specific subtype as it might be not available in the user session.

alyst · 2026-01-27T20:56:08Z

Rebased to the current devel

since it moved from struct to ctor

github-actions · 2026-01-27T21:56:04Z

ext/SEMNLOptExt/NLopt.jl


+"""
+Uses *NLopt.jl* as the optimization engine.
+Only available if *NLopt.jl* is loaded in the current Julia session!


[JuliaFormatter] _{reported by reviewdog 🐶}

Suggested change

Only available if *NLopt.jl* is loaded in the current Julia session!

!isa(inequality_constraints, NamedTuple) ||

Maximilian-Stefan-Ernst · 2026-01-27T22:12:50Z

docs/src/tutorials/backends/nlopt.md

+When [`NLopt.jl`](https://github.com/jump-dev/NLopt.jl) is loaded in the running Julia session,
+it could be used by the [`SemOptimizer`](@ref) by specifying `engine = :NLopt`
+(see [NLopt-specific options](@ref `SemOptimizer(Val(:NLopt))`)).
+Among other things, `NLopt` enables constrained optimization of the SEM models, which is
+explained in the [Constrained optimization](@ref) section.


Suggested change

When [`NLopt.jl`](https://github.com/jump-dev/NLopt.jl) is loaded in the running Julia session,

it could be used by the [`SemOptimizer`](@ref) by specifying `engine = :NLopt`

(see [NLopt-specific options](@ref `SemOptimizer(Val(:NLopt))`)).

Among other things, `NLopt` enables constrained optimization of the SEM models, which is

explained in the [Constrained optimization](@ref) section.

When [`NLopt.jl`](https://github.com/jump-dev/NLopt.jl) is loaded in the running Julia session,

it can be used by the [`SemOptimizer`](@ref) by specifying `engine = :NLopt`

(see [NLopt-specific options](@ref `SemOptimizer(Val(:NLopt))`)).

Among other things, `NLopt` enables constrained optimization of SEMs, which is

explained in the [Constrained optimization](@ref) section.

Maximilian-Stefan-Ernst · 2026-01-27T22:16:47Z

docs/src/tutorials/backends/nlopt.md

 To choose an algorithm, just pass its name without the 'NLOPT\_' prefix (for example, 'NLOPT\_LD\_SLSQP' can be used by passing `algorithm = :LD_SLSQP`).

-The README of the [julia package](https://github.com/JuliaOpt/NLopt.jl) may also be helpful, and provides a list of options:
+The README of the [*NLopt.jl*](https://github.com/JuliaOpt/NLopt.jl) may also be helpful, and provides a list of options:


Suggested change

The README of the [*NLopt.jl*](https://github.com/JuliaOpt/NLopt.jl) may also be helpful, and provides a list of options:

The README of [*NLopt.jl*](https://github.com/JuliaOpt/NLopt.jl) may also be helpful, and provides a list of options:

Maximilian-Stefan-Ernst · 2026-01-27T22:42:32Z

docs/src/tutorials/regularization/regularization.md

+You can define lasso, elastic net and other forms of regularization using [`ProximalOperators.jl`](https://github.com/JuliaFirstOrder/ProximalOperators.jl)
+and optimize the SEM model with [`ProximalAlgorithms.jl`](https://github.com/JuliaFirstOrder/ProximalAlgorithms.jl)
+that provides so-called *proximal optimization* algorithms.


Suggested change

You can define lasso, elastic net and other forms of regularization using [`ProximalOperators.jl`](https://github.com/JuliaFirstOrder/ProximalOperators.jl)

and optimize the SEM model with [`ProximalAlgorithms.jl`](https://github.com/JuliaFirstOrder/ProximalAlgorithms.jl)

that provides so-called *proximal optimization* algorithms.

You can define lasso, elastic net and other forms of regularization using [`ProximalOperators.jl`](https://github.com/JuliaFirstOrder/ProximalOperators.jl)

and optimize the SEM with so-called *proximal optimization* algorithms from [`ProximalAlgorithms.jl`](https://github.com/JuliaFirstOrder/ProximalAlgorithms.jl).

Maximilian-Stefan-Ernst · 2026-01-27T22:43:09Z

docs/src/tutorials/regularization/regularization.md

+With *ProximalAlgorithms* package loaded, it is now possible to use `:Proximal` optimization engine
+in `SemOptimizer` for estimating regularized models.


Suggested change

With *ProximalAlgorithms* package loaded, it is now possible to use `:Proximal` optimization engine

in `SemOptimizer` for estimating regularized models.

With the *ProximalAlgorithms* package loaded, it is now possible to use the `:Proximal` optimization engine

in `SemOptimizer` for estimating regularized models.

alyst · 2026-01-27T23:37:25Z

docs/src/tutorials/regularization/regularization.md

-To define a sup of separable proximal operators (i.e. no parameter is penalized twice),
-we can use [`SlicedSeparableSum`](https://juliafirstorder.github.io/ProximalOperators.jl/stable/calculus/#ProximalOperators.SlicedSeparableSum) from the `ProximalOperators` package:
+Since we apply *l1* and *l0* to the disjoint sets of parameters, this regularization could be represented as
+as sum of *separable proximal operators* (i.e. no parameter is penalized twice)


Suggested change

as sum of *separable proximal operators* (i.e. no parameter is penalized twice)

a sum of *separable proximal operators* (i.e. no parameter is penalized twice)

Maximilian-Stefan-Ernst · 2026-01-28T14:29:55Z

I opened alyst#4 to try to streamline the access of the docs a bit further - let me know what you think.

github-actions bot reviewed Jan 27, 2026

View reviewed changes

src/optimizer/abstract.jl Show resolved Hide resolved

src/optimizer/optim.jl Show resolved Hide resolved

alyst mentioned this pull request Jan 27, 2026

No-op code & docs cleanups #245

Merged

alyst and others added 13 commits January 27, 2026 12:45

Revert "fix Proximal extension"

453fcc9

This reverts commit 9729819.

Revert "fix NLopt extension"

abc2847

This reverts commit 81a4bd9.

Revert "fix exporting structs from package extensions"

56cdef1

This reverts commit f0df653.

types.jl: move SemOptimizer API into abstract.jl

421927e

NLoptResult should not be mutable

84bd7bd

SemNLOpt: use f or f => tol pair for constraints

930e0e5

It is a simple and intuitive syntax and avoids declaring new types. Also allow specifying default constraint tolerance as `constraint_tol`.

NLopt: update/simplify docs

7bd1007

use SemOptimizer(engine = :NLopt) instead of SemOptimizerNLopt() as this is a more universal scheme

Optim.md: SemOptimizerOptim => SemOptimizer

96f5f17

regulariz.md: SemOptimProx => SemOptimizer

869429b

engine(): fix signature

6d88590

optimizer_engines(): new method

3e2de1d

SemOptimizer() ctor switch to Val(E) dispatch

a560a01

It works around docstring errors during SemOptimizer{:E} parsing.

SemOptimizer: reattach docstrings to ctor

8dacd43

Instead of engine-specific subtype as it might be not available in the user session.

alyst force-pushed the refactor_semopt_api branch from fe23e46 to 8dacd43 Compare January 27, 2026 20:52

alyst marked this pull request as ready for review January 27, 2026 20:55

Alexey Stukalov added 3 commits January 27, 2026 13:50

constraints.md: cleanups

06a9a13

reg.md: cleanup

519cff1

NLopt.jl: fixup docstring

ce87dd4

since it moved from struct to ctor

github-actions bot reviewed Jan 27, 2026

View reviewed changes

Alexey Stukalov added 3 commits January 27, 2026 14:03

docs: fixup docstring switch

51121fd

tut/nlopt.md: cleanups

de9d2e8

reg.md: fixup

2b42351

Maximilian-Stefan-Ernst reviewed Jan 27, 2026

View reviewed changes

alyst commented Jan 27, 2026

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Refactor SemOptimizer API #299

Refactor SemOptimizer API #299

Uh oh!

alyst commented Jan 27, 2026

Uh oh!

Uh oh!

Uh oh!

codecov bot commented Jan 27, 2026 •

edited

Loading

Uh oh!

alyst commented Jan 27, 2026

Uh oh!

github-actions bot Jan 27, 2026

Uh oh!

Maximilian-Stefan-Ernst Jan 27, 2026

Uh oh!

Maximilian-Stefan-Ernst Jan 27, 2026

Uh oh!

Maximilian-Stefan-Ernst Jan 27, 2026

Uh oh!

Maximilian-Stefan-Ernst Jan 27, 2026

Uh oh!

alyst Jan 27, 2026

Uh oh!

Maximilian-Stefan-Ernst commented Jan 28, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

	Only available if NLopt.jl is loaded in the current Julia session!
	!isa(inequality_constraints, NamedTuple) \|\|

	The README of the [NLopt.jl](https://github.com/JuliaOpt/NLopt.jl) may also be helpful, and provides a list of options:
	The README of [NLopt.jl](https://github.com/JuliaOpt/NLopt.jl) may also be helpful, and provides a list of options:

		With ProximalAlgorithms package loaded, it is now possible to use `:Proximal` optimization engine
		in `SemOptimizer` for estimating regularized models.

	as sum of separable proximal operators (i.e. no parameter is penalized twice)
	a sum of separable proximal operators (i.e. no parameter is penalized twice)

Refactor SemOptimizer API #299

Are you sure you want to change the base?

Refactor SemOptimizer API #299

Uh oh!

Conversation

alyst commented Jan 27, 2026

Uh oh!

Uh oh!

Uh oh!

codecov bot commented Jan 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

alyst commented Jan 27, 2026

Uh oh!

github-actions bot Jan 27, 2026

Choose a reason for hiding this comment

Uh oh!

Maximilian-Stefan-Ernst Jan 27, 2026

Choose a reason for hiding this comment

Uh oh!

Maximilian-Stefan-Ernst Jan 27, 2026

Choose a reason for hiding this comment

Uh oh!

Maximilian-Stefan-Ernst Jan 27, 2026

Choose a reason for hiding this comment

Uh oh!

Maximilian-Stefan-Ernst Jan 27, 2026

Choose a reason for hiding this comment

Uh oh!

alyst Jan 27, 2026

Choose a reason for hiding this comment

Uh oh!

Maximilian-Stefan-Ernst commented Jan 28, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

codecov bot commented Jan 27, 2026 •

edited

Loading