fix backwards compat just in case

[penelopeysm]

#1331 changed the type parameters on PointwiseLogProb. While this isn't exported, I realised that it's kind of necessary to use upstream (e.g. in FlexiChains). To avoid breakage, this patch introduces a second constructor for backwards compatibility.

[github-actions]

Benchmark Report

• this PR's head: bdc140a095e2dba548c5d18d5124bcaade5a06d8
• base branch: 7d8536659fe012824bac28315788b545c291435a

Computer Information

Julia Version 1.11.9
Commit 53a02c0720c (2026-02-06 00:27 UTC)
Build Info:
  Official https://julialang.org/ release
Platform Info:
  OS: Linux (x86_64-linux-gnu)
  CPU: 4 × Intel(R) Xeon(R) Platinum 8370C CPU @ 2.80GHz
  WORD_SIZE: 64
  LLVM: libLLVM-16.0.6 (ORCJIT, icelake-server)
Threads: 1 default, 0 interactive, 1 GC (on 4 virtual cores)

Benchmark Results

┌───────────────────────┬───────┬─────────────┬────────┬───────────────────────────────┬────────────────────────────┬─────────────────────────────────┐
│                       │       │             │        │       t(eval) / t(ref)        │     t(grad) / t(eval)      │        t(grad) / t(ref)         │
│                       │       │             │        │ ─────────┬──────────┬──────── │ ───────┬─────────┬──────── │ ──────────┬───────────┬──────── │
│                 Model │   Dim │  AD Backend │ Linked │     base │  this PR │ speedup │   base │ this PR │ speedup │      base │   this PR │ speedup │
├───────────────────────┼───────┼─────────────┼────────┼──────────┼──────────┼─────────┼────────┼─────────┼─────────┼───────────┼───────────┼─────────┤
│               Dynamic │    10 │    mooncake │   true │   280.34 │   358.13 │    0.78 │   8.22 │    6.25 │    1.32 │   2304.50 │   2238.44 │    1.03 │
│                   LDA │    12 │ reversediff │   true │  2532.20 │  2549.33 │    0.99 │   2.13 │    2.50 │    0.85 │   5390.84 │   6366.59 │    0.85 │
│   Loop univariate 10k │ 10000 │    mooncake │   true │ 30664.98 │ 32098.31 │    0.96 │   6.52 │    6.75 │    0.97 │ 199898.56 │ 216508.03 │    0.92 │
├───────────────────────┼───────┼─────────────┼────────┼──────────┼──────────┼─────────┼────────┼─────────┼─────────┼───────────┼───────────┼─────────┤
│    Loop univariate 1k │  1000 │    mooncake │   true │  3123.63 │  3632.90 │    0.86 │   6.37 │    5.85 │    1.09 │  19888.91 │  21262.07 │    0.94 │
│      Multivariate 10k │ 10000 │    mooncake │   true │ 32580.20 │ 49040.03 │    0.66 │   9.69 │    7.32 │    1.32 │ 315814.09 │ 358901.64 │    0.88 │
│       Multivariate 1k │  1000 │    mooncake │   true │  3360.08 │  4551.65 │    0.74 │   9.50 │    7.12 │    1.34 │  31933.24 │  32392.16 │    0.99 │
├───────────────────────┼───────┼─────────────┼────────┼──────────┼──────────┼─────────┼────────┼─────────┼─────────┼───────────┼───────────┼─────────┤
│ Simple assume observe │     1 │ forwarddiff │  false │     0.88 │     0.85 │    1.03 │  10.42 │   11.03 │    0.95 │      9.14 │      9.36 │    0.98 │
│           Smorgasbord │   201 │ forwarddiff │  false │   944.12 │  1070.42 │    0.88 │ 151.45 │   79.05 │    1.92 │ 142991.37 │  84612.41 │    1.69 │
│           Smorgasbord │   201 │      enzyme │   true │  1290.27 │  1438.25 │    0.90 │   4.84 │    4.52 │    1.07 │   6247.26 │   6494.80 │    0.96 │
├───────────────────────┼───────┼─────────────┼────────┼──────────┼──────────┼─────────┼────────┼─────────┼─────────┼───────────┼───────────┼─────────┤
│           Smorgasbord │   201 │ forwarddiff │   true │  1287.77 │  1439.43 │    0.89 │  69.12 │   68.33 │    1.01 │  89017.32 │  98361.61 │    0.91 │
│           Smorgasbord │   201 │    mooncake │   true │  1283.29 │  1452.21 │    0.88 │   4.74 │    3.86 │    1.23 │   6081.06 │   5608.80 │    1.08 │
│           Smorgasbord │   201 │ reversediff │   true │  1291.26 │  1430.27 │    0.90 │ 127.59 │  127.15 │    1.00 │ 164756.04 │ 181856.21 │    0.91 │
├───────────────────────┼───────┼─────────────┼────────┼──────────┼──────────┼─────────┼────────┼─────────┼─────────┼───────────┼───────────┼─────────┤
│              Submodel │     1 │    mooncake │   true │     0.88 │     0.92 │    0.95 │  26.33 │   29.69 │    0.89 │     23.10 │     27.39 │    0.84 │
└───────────────────────┴───────┴─────────────┴────────┴──────────┴──────────┴─────────┴────────┴─────────┴─────────┴───────────┴───────────┴─────────┘

[codecov]

Codecov Report

❌ Patch coverage is 0% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 82.26%. Comparing base (81fe531) to head (bdc140a).
⚠️ Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
src/accumulators/pointwise_logdensities.jl	0.00%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1375      +/-   ##
==========================================
- Coverage   82.30%   82.26%   -0.05%     
==========================================
  Files          49       49              
  Lines        3510     3512       +2     
==========================================
  Hits         2889     2889              
- Misses        621      623       +2     

☔ 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.