Feature/crispino nontransitive - #83
Merged
Merged
Conversation
There was a problem hiding this comment.
Pull request overview
This PR adds support for a Bernoulli error model for pairwise-preference data (to handle non-transitive comparisons), introduces a Beta prior for the error probability epsilon, and returns/traces epsilon from the SMC run.
Changes:
- Add
error_modeloption ("none"/"bernoulli") and thread it through the C++ SMC pipeline. - Introduce
epsilonas a static parameter with Beta prior hyperparameterskappa_1/kappa_2, and surfaceepsilon+epsilon_tracesin the results. - Add a test for non-transitive pairwise preferences under the Bernoulli error model, plus NEWS/docs/version bumps.
Reviewed changes
Copilot reviewed 20 out of 20 changed files in this pull request and generated 7 comments.
Show a summary per file
| File | Description |
|---|---|
| tests/testthat/test-compute_sequentially_bernoulli.R | Adds a test for Bernoulli error model on non-transitive pairwise preferences. |
| src/sample_latent_rankings.h | Extends latent ranking proposal API with error_model. |
| src/sample_latent_rankings.cpp | Implements Bernoulli-path latent ranking proposals for pairwise data. |
| src/run_smc.cpp | Returns epsilon and epsilon_traces in the model output. |
| src/rejuvenate.cpp | Updates rejuvenation to carry epsilon and includes a Gibbs-style epsilon update. |
| src/prior.h | Adds kappa_1/kappa_2 to the prior struct. |
| src/prior.cpp | Reads kappa_1/kappa_2 from the R prior list. |
| src/particle.h | Adds epsilon to StaticParameters and updates constructors. |
| src/particle.cpp | Initializes epsilon, passes error_model to proposals, and adds Bernoulli likelihood contribution for pairwise data. |
| src/parameter_tracer.h | Adds epsilon_traces. |
| src/parameter_tracer.cpp | Records epsilon over time when tracing is enabled. |
| src/options.h | Adds error_model option field. |
| src/options.cpp | Reads error_model from R options list. |
| R/set_smc_options.R | Adds error_model argument and docs. |
| R/set_hyperparameters.R | Adds kappa_1/kappa_2 hyperparameters and docs. |
| NEWS.md | Documents the new Bernoulli error model and epsilon tracking. |
| man/set_smc_options.Rd | Updates Rd for new error_model option. |
| man/set_hyperparameters.Rd | Updates Rd for kappa_1/kappa_2. |
| DESCRIPTION | Bumps package version to development version 0.3.0.9000. |
| CRAN-SUBMISSION | Adds CRAN submission metadata file. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
Comment on lines
+9
to
+10
| kappa_1 { input_prior["kappa_1"] }, | ||
| kappa_2 { input_prior["kappa_2"] } {} |
Comment on lines
+19
to
+22
| epsilon = Rcpp::rbeta(1, prior.kappa_1, prior.kappa_2)[0]; | ||
| if (epsilon >= 0.5) { | ||
| epsilon = 0.4999; // Simple way to handle the truncation for the initial value | ||
| } |
Comment on lines
+164
to
+167
| double epsilon_prime = 1.0; | ||
| while (epsilon_prime >= 0.5) { | ||
| epsilon_prime = Rcpp::rbeta(1, prior.kappa_1 + total_a, prior.kappa_2 + total_b)[0]; | ||
| } |
Comment on lines
+148
to
+151
| pairwise_tp new_data = pp->timeseries[t]; | ||
| for (auto ndit = new_data.begin(); ndit != new_data.end(); ++ndit) { | ||
| total_p += ndit->second.size(); | ||
| for (auto pair : ndit->second) { |
Comment on lines
+138
to
+144
| umat sort_matrix = new_sort_matrices[ndit->first]; | ||
| int random_index = Rcpp::sample(sort_matrix.n_cols, 1, false)[0] - 1; | ||
|
|
||
| proposal.proposal.col(proposal_index++) = sort_matrix.col(random_index); | ||
| proposal.log_probability = join_vert( | ||
| proposal.log_probability, vec{-log(new_sort_counts[ndit->first])} | ||
| ); | ||
| proposal.proposal.col(proposal_index++) = sort_matrix.col(random_index); | ||
| proposal.log_probability = join_vert( | ||
| proposal.log_probability, vec{-log(new_sort_counts[ndit->first])} | ||
| ); |
Comment on lines
+22
to
+38
| # Run compute_sequentially | ||
| mod <- compute_sequentially( | ||
| df, | ||
| topological_sorts = top_sorts, | ||
| smc_options = set_smc_options( | ||
| n_particles = 100, | ||
| n_particle_filters = 5, | ||
| error_model = "bernoulli", | ||
| use_backward_simulation = TRUE | ||
| ), | ||
| hyperparameters = set_hyperparameters(n_items = 3) | ||
| ) | ||
|
|
||
| # Assertions | ||
| expect_true(!is.null(mod$epsilon)) | ||
| expect_true(mean(mod$epsilon) > 0) | ||
| expect_true(mean(mod$epsilon) < 0.5) |
| @@ -0,0 +1,3 @@ | |||
| Version: 0.3.0 | |||
…, fixed ranking conversion bug, and test fixes
Owner
Author
|
@copilot can you tell me if the updated commit passes your review? |
Contributor
2aa3074 addresses the specific review comments I raised earlier, but I wouldn’t mark it as fully passing yet because the current R-CMD-check run is failing on ubuntu-latest (release) and ubuntu-latest (oldrel-1). Both failures are in |
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
No description provided.