Repository navigation
[FEAT] IncrementalAdditive Cleanup for Chapter 16 Part 1 - #1419
henrydingliu wants to merge 14 commits into
Conversation
Pyright Type CompletenessView the full Project (full
Other symbols referenced but not exported by
Symbols without documentation:
Patch (exported symbols added or changed by this PR): 33.3% fully typed (2 / 6); 1 no longer exported
Patch symbol details
|
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #1419 +/- ##
==========================================
+ Coverage 91.79% 92.95% +1.16%
==========================================
Files 96 102 +6
Lines 5494 6092 +598
Branches 707 771 +64
==========================================
+ Hits 5043 5663 +620
+ Misses 327 310 -17
+ Partials 124 119 -5
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
|
@genedan just assigned this to you as i'd like this to land so i can continue working on chapter 16. |
|
I'm working my way through the open PRs bottom up, so this one is next. |
There was a problem hiding this comment.
Cursor Bugbot has reviewed your changes and found 1 potential issue.
❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.
Reviewed by Cursor Bugbot for commit d7a983e. Configure here.
|
I'm doing a pre-thanksgiving clean-up. @genedan do you have bandwidth to review this? |

Summary of Changes
More clean-up and bug fix as I work through Chapter 16
zeta_-related properties into IncrementalMixin, matching DisposalMixincum_to_incrforcum_zeta_, w/ new test to guard future drift; also improved readability of the surrounding code_param_propertyacross modules to condense the universal properties toDevelopmentBaseand keep module-specific properties at module levelRelated GitHub Issue(s)
#588
#377
AI/LLM Usage
AI did some autofill of docstrings, which I then validated edited
Additional Context for Reviewers
zeta_ still shows the wrong development labels. will address after #1304
Submitter's Checklist
[FIX],[FEAT],[DOCS],[TST],[CHORE], or[BRK]).Reviewer's Checklist
Note
Medium Risk
Changes triangle incr/cum conversion and how fitted development parameters are tagged, which can affect incremental additive and pattern-based workflows; mitigated by targeted tests.
Overview
Incremental additive (
zeta_) handling is aligned with the disposal mixin pattern: a newIncrementalMixinonTriangleBaseexposeszeta_/cum_zeta_with setters that stamp pattern flags (is_pattern,is_additive,is_cumulative), stores values on_zeta_, andhas_zetanow keys off that private attribute.cum_zeta_is removed from the sharedCommonmixin in favor of the incremental mixin API.Parameter triangles are refactored so
DevelopmentBase._param_propertyonly sets shared pattern metadata;Development,IncrementalAdditive, and disposal estimators override it for type-specific flags (multiplicative LDF vs additivezeta_vs disposal rates).Development.fitno longer stacks regression outputs into one array with an index—each statistic gets its own_param_propertycall.incr_to_cum/cum_to_incronTriangleare rewritten with explicit branches for multiplicative LDF patterns, additive factor patterns (is_additive), normal increments, and disposal rates, plus a dedicatedis_additiveproperty. This fixes incorrect incremental conversion for cumulative additive (cum_zeta_) patterns.Tests cover setting
cum_zeta_, missingzeta_errors, and updatedhas_zeta/_param_propertycall sites.Reviewed by Cursor Bugbot for commit 4b3e652. Bugbot is set up for automated code reviews on this repo. Configure here.