Implement values_as_in_model using an accumulator

[mhauru]

Also adds a relevant test.

[github-actions]

Benchmark Report for Commit c2a8e63

Computer Information

Julia Version 1.11.5
Commit 760b2e5b739 (2025-04-14 06:53 UTC)
Build Info:
  Official https://julialang.org/ release
Platform Info:
  OS: Linux (x86_64-linux-gnu)
  CPU: 4 × AMD EPYC 7763 64-Core Processor
  WORD_SIZE: 64
  LLVM: libLLVM-16.0.6 (ORCJIT, znver3)
Threads: 1 default, 0 interactive, 1 GC (on 4 virtual cores)

Benchmark Results

|                 Model | Dimension |  AD Backend |      VarInfo Type | Linked | Eval Time / Ref Time | AD Time / Eval Time |
|-----------------------|-----------|-------------|-------------------|--------|----------------------|---------------------|
| Simple assume observe |         1 | forwarddiff |             typed |  false |                  8.9 |                 1.7 |
|           Smorgasbord |       201 | forwarddiff |             typed |  false |                703.4 |                36.9 |
|           Smorgasbord |       201 | forwarddiff | simple_namedtuple |   true |                432.5 |                47.8 |
|           Smorgasbord |       201 | forwarddiff |           untyped |   true |               1287.6 |                27.0 |
|           Smorgasbord |       201 | forwarddiff |       simple_dict |   true |               8273.4 |                20.1 |
|           Smorgasbord |       201 | reversediff |             typed |   true |               1576.4 |                26.5 |
|           Smorgasbord |       201 |    mooncake |             typed |   true |               1034.9 |                 4.9 |
|    Loop univariate 1k |      1000 |    mooncake |             typed |   true |               6045.1 |                 3.9 |
|       Multivariate 1k |      1000 |    mooncake |             typed |   true |               1040.3 |                 8.8 |
|   Loop univariate 10k |     10000 |    mooncake |             typed |   true |              67978.2 |                 3.5 |
|      Multivariate 10k |     10000 |    mooncake |             typed |   true |               9084.4 |                 9.8 |
|               Dynamic |        10 |    mooncake |             typed |   true |                144.2 |                12.3 |
|              Submodel |         1 |    mooncake |             typed |   true |                 13.1 |                 6.3 |
|                   LDA |        12 | reversediff |             typed |   true |               1532.4 |                 2.7 |

[codecov]

Codecov Report

Attention: Patch coverage is 75.00000% with 8 lines in your changes missing coverage. Please review.

Please upload report for BASE (breaking@326d7ed). Learn more about missing BASE report.

Files with missing lines	Patch %	Lines
src/compiler.jl	40.00%	6 Missing ⚠️
src/values_as_in_model.jl	90.90%	2 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##             breaking     #908   +/-   ##
===========================================
  Coverage            ?   84.67%           
===========================================
  Files               ?       37           
  Lines               ?     4040           
  Branches            ?        0           
===========================================
  Hits                ?     3421           
  Misses              ?      619           
  Partials            ?        0           

☔ 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.

[penelopeysm]

Broader comment, since we discussed the possibility of taking VAIMAcc further just now (I opened an issue to track that too TuringLang/Turing.jl#2545). Is ValuesAsInModelAccumulator the best name, or is there something more suitable? It is a bit of a mouthful and I feel like we have a chance to rename it here

[mhauru]

I like that it has the same name as the only function that uses it, and renaming the function would be breaking.

[penelopeysm]

The docs failure is because of GHA perms, I'll fix that on main