Update to v0.14.0

[networmix]

• Restructured Monte Carlo results: baseline is returned separately, and results now contains deduplicated failure patterns with occurrence_count.
• Removed baseline parameter from Monte Carlo APIs; baseline now runs implicitly.
• Added FlowIterationResult.occurrence_count to track how many iterations produced each failure pattern.

---

Note

Monte Carlo workflows now run a separate baseline automatically, dedupe failure iterations with occurrence_count, add optional failure_trace, remove the baseline parameter, and update steps, docs, scenarios, and tests accordingly (v0.14.0).

• Core/Execution:
  • FailureManager: always runs a separate baseline, deduplicates failure patterns, tracks occurrence_count, optional failure_trace; parallel/serial runners simplified.
• Workflow Steps:
  • MaxFlow and TrafficMatrixPlacement: removed baseline param; accept seed; write data.baseline plus deduped data.flow_results; logging/metadata tweaks.
• Results Schema:
  • FlowIterationResult: add occurrence_count and failure_trace; to_dict/export updated.
  • Result shape: baseline separate; flow_results = unique patterns; metadata includes unique_patterns.
• Failure Policies:
  • apply_failures(...) accepts failure_trace and records mode/rule selections and expansions.
• Docs/DSL:
  • References/examples updated to new schema (baseline implicit, no baseline: field in YAML); added params (seeds, include/exclude lists); examples count increased.
• Scenarios/Tests:
  • All scenarios and tests updated to new result shape and removed baseline parameter; new tests for tracing, dedup, weighting.
• Version:
  • Bump to 0.14.0; changelog added with breaking changes.

^{Written by Cursor Bugbot for commit c399e92. This will update automatically on new commits. Configure here.}

[Copilot]

Pull request overview

This PR introduces version 0.14.0 with breaking changes to the Monte Carlo analysis API. The main purpose is to restructure how Monte Carlo results are returned: baseline results are now separated from failure iterations, and failure results are deduplicated with an occurrence_count field to track pattern frequency.

Key Changes

• Removed the baseline parameter from Monte Carlo APIs; baseline is now always run implicitly and returned separately
• Restructured results to separate baseline from results array, with deduplicated failure patterns tracked by occurrence_count
• Added FlowIterationResult.failure_trace to capture detailed failure policy execution trace when store_failure_patterns=True

Reviewed changes

Copilot reviewed 43 out of 43 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
ngraph/exec/failure/manager.py	Core refactoring: baseline always runs first, results deduplicated with occurrence_count tracking, failure_trace support
ngraph/results/flow.py	Added failure_trace and occurrence_count fields to FlowIterationResult with validation
ngraph/model/failure/policy.py	Extended apply_failures() to populate failure_trace dict when requested
ngraph/workflow/max_flow_step.py	Removed baseline parameter, updated to handle separate baseline result
ngraph/workflow/traffic_matrix_placement_step.py	Removed baseline parameter, updated to handle separate baseline result
tests/exec/failure/test_manager.py	Comprehensive test updates for new API, added tests for occurrence_count weighting
tests/exec/failure/test_manager_integration.py	Updated tests for deduplicated results and failure_trace
tests/model/failure/test_failure_trace.py	New test file for failure_trace functionality (378 lines)
tests/workflow/*.py	Updated all workflow tests to remove baseline parameter and handle new result structure
scenarios/*.yaml	Removed baseline parameter from all scenario YAML files
docs/*	Updated documentation to reflect API changes
CHANGELOG.md	Documented breaking changes for v0.14.0
pyproject.toml, ngraph/_version.py	Version bump to 0.14.0

Comments suppressed due to low confidence (1)

ngraph/exec/failure/manager.py:590

• The metadata field 'unique_patterns' is computed from 'num_unique_tasks', which represents unique failure patterns. However, when iterations=0 (no failures), this will be 0. The metadata should accurately reflect that unique_patterns refers to failure patterns, not including the baseline. Consider explicitly documenting this behavior or verifying that num_unique_tasks is correctly set when there are no failure iterations.

            "metadata": {
                "iterations": iterations,
                "parallelism": parallelism,
                "analysis_function": func_name,
                "policy_name": self.policy_name,
                "execution_time": elapsed_time,
                "unique_patterns": num_unique_tasks,

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

The validation error message should clarify that occurrence_count must be a positive integer (greater than or equal to 1), not just any positive value. The current message "occurrence_count must be a positive int" is accurate but could be more explicit about the minimum value of 1.

Suggested change

	"FlowIterationResult.occurrence_count must be a positive int: %r",
	self.occurrence_count,
	)
	raise ValueError("occurrence_count must be a positive int")
	"FlowIterationResult.occurrence_count must be a positive integer (>= 1): %r",
	self.occurrence_count,
	)
	raise ValueError(
	"occurrence_count must be a positive integer (>= 1)"
	)