Modify inference for predictions functionality

[albertpod]

This PR aims at bringing the prediction functionality into RxInfer.jl.

[albertpod]

Short summary of the modifications:

• The functions make_actor have been modified to accept AbstractVariable instead of RandomVariable.
• The InferenceResult struct has been expanded to include a new field named predictions.
• New functions __inference_check_dataismissing, __inference_fill_predictions have been added to handle predictions.
• The inference function has been updated to handle predictions.
• The data argument is now optional if predictvars are specified.
• New logic has been added to handle prediction variables.
• Subscription logic has been updated to handle both types of variables.
• Several error checks and warnings have been added or modified to ensure that the data and prediction variables are correctly specified and handled.

[albertpod]

NOTE: The prediction functionality doesn't allow free energy computation at the moment. I have error when free_energy and predictions are both present. The research question remains: how do we compute FE on non-terminated graph (ping @ismailsenoz )

[review-notebook-app[bot]]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[codecov[bot]]

Codecov Report

Patch coverage: 76.71% and project coverage change: -0.29% :warning:

Comparison is base (2ba4dea) 81.32% compared to head (c309034) 81.03%.

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #51 +/- ## ========================================== - Coverage 81.32% 81.03% -0.29% ========================================== Files 11 11 Lines 1210 1271 +61 ========================================== + Hits 984 1030 +46 - Misses 226 241 +15 ``` | [Files Changed](https://app.codecov.io/gh/biaslab/RxInfer.jl/pull/51?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=biaslab) | Coverage Δ | | |---|---|---| | [src/inference.jl](https://app.codecov.io/gh/biaslab/RxInfer.jl/pull/51?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=biaslab#diff-c3JjL2luZmVyZW5jZS5qbA==) | `75.77% <74.60%> (+0.07%)` | :arrow_up: | | [src/model.jl](https://app.codecov.io/gh/biaslab/RxInfer.jl/pull/51?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=biaslab#diff-c3JjL21vZGVsLmps) | `85.85% <90.00%> (+0.14%)` | :arrow_up: | ... and [1 file with indirect coverage changes](https://app.codecov.io/gh/biaslab/RxInfer.jl/pull/51/indirect-changes?src=pr&el=tree-more&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=biaslab)

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.