Task1 lpdid scaffold

CHANGELOG.md

@@ -7,6 +7,16 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+### Added
+- **LPDiD estimator** (`LPDiD`) for the absorbing-treatment main path of Stata `lpdid`
+  - Event-study and pooled pre/post effects with clean controls or never-treated controls
+  - `rw` and `nocomp` support with official Stata parity tests on bundled example data
+  - `pmd(max|k)`, covariates, `ylags`, `dylags`, and additional absorbed fixed effects
+  - Regression-adjustment parity for `rw + controls/absorb/ylags/dylags` on focused Stata-validated samples
+
+### Changed
+- Move Local Projections DiD from roadmap item to implemented estimator in public documentation
+
 ## [2.8.0] - 2026-03-31
 
 ### Added

README.md

@@ -85,6 +85,7 @@ Detailed guide: [`docs/llms-practitioner.txt`](docs/llms-practitioner.txt)
 - **Panel data support**: Two-way fixed effects estimator for panel designs
 - **Multi-period analysis**: Event-study style DiD with period-specific treatment effects
 - **Staggered adoption**: Callaway-Sant'Anna (2021), Sun-Abraham (2021), Borusyak-Jaravel-Spiess (2024) imputation, Two-Stage DiD (Gardner 2022), Stacked DiD (Wing, Freedman & Hollingsworth 2024), and Efficient DiD (Chen, Sant'Anna & Xie 2025) estimators for heterogeneous treatment timing
+- **Local Projections DiD**: `LPDiD` for absorbing-treatment event studies and pooled effects, with clean controls, `rw`, `nocomp`, `pmd`, covariates, lags, and additional absorbed fixed effects
 - **Triple Difference (DDD)**: Ortiz-Villavicencio & Sant'Anna (2025) estimators with proper covariate handling
 - **Synthetic DiD**: Combined DiD with synthetic control for improved robustness
 - **Triply Robust Panel (TROP)**: Factor-adjusted DiD with synthetic weights (Athey et al. 2025)
@@ -117,9 +118,42 @@ All estimators have short aliases for convenience:
 | `Stacked` | `StackedDiD` | Stacked DiD |
 | `Bacon` | `BaconDecomposition` | Goodman-Bacon decomposition |
 | `EDiD` | `EfficientDiD` | Efficient DiD |
+| `LPDiD` | `LPDiD` | Local Projections DiD |
 
 `TROP` already uses its short canonical name and needs no alias.
 
+## Local Projections DiD
+
+`LPDiD` ports the absorbing-treatment main path of Stata's `lpdid` command into
+the `diff-diff` estimator interface. It supports dynamic event-study effects and
+pooled pre/post effects with clean controls or never-treated controls, optional
+reweighting (`rw`), common-composition samples (`nocomp`), pre-mean
+differencing (`pmd`), covariates, outcome lags, first-difference lags, and
+additional absorbed fixed effects.
+
+```python
+from diff_diff import LPDiD
+
+res = LPDiD(
+    pre_window=5,
+    post_window=10,
+    reweight=True,
+    no_composition=True,
+).fit(
+    data,
+    outcome="Y",
+    unit="unit",
+    time="time",
+    treatment="treat",
+)
+
+print(res.event_study.head())
+print(res.pooled)
+```
+
+The current implementation is validated against official Stata `lpdid`
+absorbing-treatment examples and option paths used in the test suite.
+
 ## Tutorials
 
 We provide Jupyter notebook tutorials in `docs/tutorials/`:
@@ -140,6 +174,7 @@ We provide Jupyter notebook tutorials in `docs/tutorials/`:
 | `12_two_stage_did.ipynb` | Two-Stage DiD (Gardner 2022), GMM sandwich variance, per-observation effects |
 | `13_stacked_did.ipynb` | Stacked DiD (Wing et al. 2024), Q-weights, sub-experiment inspection, trimming, clean control definitions |
 | `15_efficient_did.ipynb` | Efficient DiD (Chen et al. 2025), optimal weighting, PT-All vs PT-Post, efficiency gains, bootstrap inference |
+| `16_lpdid.ipynb` | LPDiD for absorbing treatments, official Stata `lpdid` example workflow, `rw`, `nocomp`, and event-study plotting |
 
 ## Data Preparation
 

ROADMAP.md

@@ -10,7 +10,7 @@ For past changes and release history, see [CHANGELOG.md](CHANGELOG.md).
 
 diff-diff v2.7.5 is a **production-ready** DiD library with feature parity with R's `did` + `HonestDiD` + `synthdid` ecosystem for core DiD analysis, plus **unique survey support** — design-based variance estimation (Taylor linearization, replicate weights) integrated across all estimators. No R or Python package offers this combination:
 
-- **Core estimators**: Basic DiD, TWFE, MultiPeriod, Callaway-Sant'Anna, Sun-Abraham, Borusyak-Jaravel-Spiess Imputation, Synthetic DiD, Triple Difference (DDD), TROP, Two-Stage DiD (Gardner 2022), Stacked DiD (Wing et al. 2024), Continuous DiD (Callaway, Goodman-Bacon & Sant'Anna 2024)
+- **Core estimators**: Basic DiD, TWFE, MultiPeriod, Callaway-Sant'Anna, Sun-Abraham, Borusyak-Jaravel-Spiess Imputation, Synthetic DiD, LPDiD, Triple Difference (DDD), TROP, Two-Stage DiD (Gardner 2022), Stacked DiD (Wing et al. 2024), Continuous DiD (Callaway, Goodman-Bacon & Sant'Anna 2024)
 - **Valid inference**: Robust SEs, cluster SEs, wild bootstrap, multiplier bootstrap, placebo-based variance
 - **Assumption diagnostics**: Parallel trends tests, placebo tests, Goodman-Bacon decomposition
 - **Sensitivity analysis**: Honest DiD (Rambachan-Roth), Pre-trends power analysis (Roth 2022)
@@ -50,6 +50,19 @@ Extend the existing `TripleDifference` estimator to handle staggered adoption se
 
 **Reference**: [Ortiz-Villavicencio & Sant'Anna (2025)](https://arxiv.org/abs/2505.09942). "Better Understanding Triple Differences Estimators." *Working Paper*. R package: `triplediff`.
 
+### Local Projections DiD
+
+The absorbing-treatment main path is now implemented as `LPDiD`, including
+event-study and pooled effects, clean vs never-treated controls, `rw`,
+`nocomp`, `pmd`, covariates, lags, and additional absorbed fixed effects with
+Stata-validated parity coverage.
+
+Remaining work is now narrower and more specialized:
+
+- nonabsorbing-treatment support
+- bootstrap inference parity for LPDiD-specific paths
+- broader documentation examples
+
 ---
 
 ## Medium-Term Enhancements
@@ -70,16 +83,6 @@ Handles treatment that switches on and off (reversible treatments), unlike most
 
 **Reference**: [de Chaisemartin & D'Haultfœuille (2020, 2024)](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3980758). *American Economic Review*.
 
-### Local Projections DiD
-
-Implements local projections for dynamic treatment effects. Doesn't require specifying full dynamic structure.
-
-- Flexible impulse response estimation
-- Robust to misspecification of dynamics
-- Natural handling of anticipation effects
-
-**Reference**: Dube, Girardi, Jordà, and Taylor (2023).
-
 ### Nonlinear DiD
 
 For outcomes where linear models are inappropriate (binary, count, bounded).

diff_diff/__init__.py

@@ -28,6 +28,9 @@
     Comparison2x2,
     bacon_decompose,
 )
+from diff_diff.lpdid import (
+    LPDiD,
+)
 from diff_diff.diagnostics import (
     PlaceboTestResults,
     leave_one_out_test,
@@ -102,6 +105,9 @@
     PeriodEffect,
     SyntheticDiDResults,
 )
+from diff_diff.lpdid_results import (
+    LPDiDResults,
+)
 from diff_diff.survey import (
     DEFFDiagnostics,
     SurveyDesign,
@@ -224,6 +230,7 @@
     "TripleDifference",
     "TROP",
     "StackedDiD",
+    "LPDiD",
     # Estimator aliases (short names)
     "DiD",
     "TWFE",
@@ -269,6 +276,7 @@
     "trop",
     "StackedDiDResults",
     "stacked_did",
+    "LPDiDResults",
     # EfficientDiD
     "EfficientDiD",
     "EfficientDiDResults",