StataFlow 0.1.2__tar.gz → 0.1.4__tar.gz

stataflow-0.1.4/PKG-INFO

@@ -0,0 +1,202 @@
+Metadata-Version: 2.4
+Name: StataFlow
+Version: 0.1.4
+Summary: StataFlow: A Python econometrics toolkit aligned with Stata 17
+Author-email: Zhenhao Fu <zhenhaofu2001@gmail.com>
+License: MIT
+Project-URL: Homepage, https://github.com/ZhenHaoFu810/StataFlow
+Project-URL: Repository, https://github.com/ZhenHaoFu810/StataFlow
+Project-URL: Issues, https://github.com/ZhenHaoFu810/StataFlow/issues
+Keywords: econometrics,stata,regression,fixed-effects,panel-data
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Scientific/Engineering :: Mathematics
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: numpy>=1.24
+Requires-Dist: pandas>=2.0
+Requires-Dist: scipy>=1.10
+Requires-Dist: pyyaml>=6.0
+Provides-Extra: dev
+Requires-Dist: pytest>=7.0; extra == "dev"
+Requires-Dist: pytest-cov>=4.0; extra == "dev"
+Dynamic: license-file
+
+# StataFlow
+
+StataFlow (`stataflow`) is a Python econometrics toolkit that reproduces Stata 17 estimation results with high precision. It provides both a **Stata-compatible command layer** (for researchers migrating from Stata) and a **native Python estimator layer** (for advanced users who want direct control).
+
+## What you can do today
+
+- Run Stata-style commands in Python: `regress`, `reghdfe`, `ivregress 2sls`, `logit`, `ppmlhdfe`, `did_imputation`, `csdid`, `rdrobust`, and more.
+- Obtain coefficients, standard errors, t/z-statistics, p-values, and confidence intervals that are field-level verified against Stata 17.
+- Work with high-dimensional fixed effects (HDFE), IV/2SLS, binary/count models, and DID/event-study estimators.
+- Use Stata-style factor-variable syntax (`i.group##c.post`, `c.x1#c.x2`, `x1##x2`) and space-separated absorb strings directly in wrapper commands. Bare variables inside `#` / `##` are treated as continuous, matching common Stata usage.
+
+## What is not yet supported
+
+- **Multi-way clustering** — only single-cluster robust inference is available.
+- **Direct post-estimation on wrapper returns** — the `compat.stata` wrappers return `ResultSchema` result objects. `predict` and `margins` are available on the core estimator layer only.
+- **Full command surfaces for community commands** — `reghdfe`, `ivreghdfe`, `ppmlhdfe`, `did_imputation`, `eventstudyinteract`, `csdid`, and `rdrobust` are implemented as **verified high-frequency subsets**, not complete Stata command reproductions. Unsupported options are explicitly rejected rather than silently ignored.
+
+### Completeness legend
+
+- **Stable** — synthetic + real-data dual-run verified; core API is unlikely to change.
+- **Alpha** — high-frequency paths are implemented and verified, but the command surface is still a subset of the full Stata community command.
+- **Alpha — Partial** — a verifiable implementation exists, but large functional areas are still missing (e.g., fuzzy RD for `rdrobust`, multi-way clustering).
+
+See the [Command Support Matrix](./docs/command-support-matrix/README.md) for the per-command detailed status.
+
+---
+
+## Installation
+
+```bash
+pip install StataFlow
+```
+
+Requirements: Python 3.10+, NumPy, pandas, SciPy.
+
+For development (editable install from source):
+
+```bash
+git clone https://github.com/ZhenHaoFu810/StataFlow.git
+cd StataFlow
+pip install -e .
+```
+
+---
+
+## Quick start
+
+### Stata-compatible command layer (recommended)
+
+All `compat.stata` wrappers return a `ResultSchema` object with coefficients, standard errors, and fit statistics. They do **not** expose `.predict()` or `.margins()` directly—use the core estimator layer below for post-estimation.
+
+```python
+import pandas as pd
+from stataflow.compat.stata import regress, reghdfe, ivregress_2sls, logit
+
+# OLS with robust standard errors
+result = regress(df, y="wage", x=["edu", "exper"], vce="robust")
+
+# High-dimensional fixed effects (reghdfe)
+result = reghdfe(
+    df, y="wage", x=["edu", "exper"],
+    absorb="firm_id year_id", vce="cluster", cluster="industry"
+)
+
+# Factor-variable syntax in HDFE
+result = reghdfe(
+    df, y="wage", x=["i.industry##c.post"], absorb="firm_id year_id"
+)
+
+# 2SLS
+result = ivregress_2sls(
+    df, y="lwage", x_exog=["edu"], x_endog=["exper"],
+    instruments=["age", "kidslt6"], vce="robust"
+)
+
+# Logit
+result = logit(df, y="inlf", x=["nwifeinc", "educ", "exper"])
+```
+
+For runnable examples, see the [`examples/`](./examples/) directory:
+- [`examples/demo_regress.py`](./examples/demo_regress.py)
+- [`examples/demo_reghdfe.py`](./examples/demo_reghdfe.py)
+- [`examples/demo_ppmlhdfe.py`](./examples/demo_ppmlhdfe.py)
+- [`examples/demo_ivregress_2sls.py`](./examples/demo_ivregress_2sls.py)
+
+### Native Python estimator layer (advanced)
+
+```python
+from stataflow import OLS, FixedEffectsOLS, AbsorbingOLS, Logit, IV2SLS
+
+model = OLS(data=df, y="wage", x=["edu", "exper"])
+result = model.fit(vce="robust")
+```
+
+---
+
+## Supported commands
+
+| Command | Python entry | Core capabilities |
+|---------|--------------|-------------------|
+| `regress` | `stataflow.compat.stata.regress` | OLS, robust, cluster, aweight |
+| `xtreg, fe` | `stataflow.compat.stata.xtreg_fe` | Fixed effects (within), cluster |
+| `areg` | `stataflow.compat.stata.areg` | Single absorb variable FE |
+| `reghdfe` | `stataflow.compat.stata.reghdfe` | 1-2 group HDFE, cluster, singleton drop |
+| `ivregress 2sls` | `stataflow.compat.stata.ivregress_2sls` | 2SLS, robust, cluster |
+| `ivreghdfe` | `stataflow.compat.stata.ivreghdfe` | IV + 1-2 group HDFE, cluster |
+| `logit` | `stataflow.compat.stata.logit` | MLE, robust, cluster |
+| `probit` | `stataflow.compat.stata.probit` | MLE, robust, cluster |
+| `poisson` | `stataflow.compat.stata.poisson` | MLE, robust, cluster |
+| `ppmlhdfe` | `stataflow.compat.stata.ppmlhdfe` | PPML + 1-2 group HDFE |
+| `did_imputation` | `stataflow.compat.stata.did_imputation` | BJS DID imputation |
+| `eventstudyinteract` | `stataflow.compat.stata.eventstudyinteract` | Sun & Abraham IW estimator |
+| `csdid` | `stataflow.compat.stata.csdid` | Callaway-Sant'Anna DID (`method="reg"` only) |
+| `rdrobust` | `stataflow.compat.stata.rdrobust` | Sharp RD local polynomial (`bwselect="mserd"`, `covs`) |
+
+Full details: [`docs/command-support-matrix/README.md`](./docs/command-support-matrix/README.md)
+
+---
+
+## Validation philosophy
+
+Every public command is validated with **two lines of evidence**:
+
+1. **Synthetic / controlled cases** — formula, degrees of freedom, sample screening, edge cases.
+2. **Real public datasets** — field-level comparison against Stata 17 on openly available economic/financial data.
+
+A command is considered "done" only when both lines pass and the source-to-Python mapping is documented. We do not accept "statistical equivalence" without explicit mathematical or source-code justification.
+
+Public evidence and results are available in `research/results/validation/`.
+
+### Running tests
+
+```bash
+# Unit and integration tests (fast)
+pytest tests/ -v --ignore=tests/golden/
+
+# Golden dual-run tests (require Stata 17)
+pytest tests/golden/ -v
+```
+
+---
+
+## Project structure
+
+- **`src/stataflow/estimators/`** — Core Python estimators (`OLS`, `AbsorbingOLS`, `Logit`, `PPMLHDFE`, `DIDImputation`, etc.)
+- **`src/stataflow/compat/stata/`** — Stata command wrappers (`regress()`, `reghdfe()`, `ivregress_2sls()`, etc.)
+- **`docs/command-support-matrix/`** — Per-command support matrices
+- **`examples/`** — Runnable demonstration scripts
+- **`tests/`** — Unit and integration tests
+
+---
+
+## Default target version
+
+**Stata 17**
+
+---
+
+## Documentation
+
+- [User guide](./docs/USER_GUIDE.md)
+- [Chinese user guide](./docs/USER_GUIDE.zh-CN.md)
+- [Cookbook](./docs/cookbook.md)
+- [Chinese cookbook](./docs/cookbook.zh-CN.md)
+- [Command support matrices](./docs/command-support-matrix/README.md)
+
+---
+
+## Governance
+
+- **Codex** — project goals, architecture, review gates, and statistical-dispute arbitration.
+- **Claude Code** — implementation, testing, and evidence backfill.

stataflow-0.1.4/README.md

@@ -0,0 +1,172 @@
+# StataFlow
+
+StataFlow (`stataflow`) is a Python econometrics toolkit that reproduces Stata 17 estimation results with high precision. It provides both a **Stata-compatible command layer** (for researchers migrating from Stata) and a **native Python estimator layer** (for advanced users who want direct control).
+
+## What you can do today
+
+- Run Stata-style commands in Python: `regress`, `reghdfe`, `ivregress 2sls`, `logit`, `ppmlhdfe`, `did_imputation`, `csdid`, `rdrobust`, and more.
+- Obtain coefficients, standard errors, t/z-statistics, p-values, and confidence intervals that are field-level verified against Stata 17.
+- Work with high-dimensional fixed effects (HDFE), IV/2SLS, binary/count models, and DID/event-study estimators.
+- Use Stata-style factor-variable syntax (`i.group##c.post`, `c.x1#c.x2`, `x1##x2`) and space-separated absorb strings directly in wrapper commands. Bare variables inside `#` / `##` are treated as continuous, matching common Stata usage.
+
+## What is not yet supported
+
+- **Multi-way clustering** — only single-cluster robust inference is available.
+- **Direct post-estimation on wrapper returns** — the `compat.stata` wrappers return `ResultSchema` result objects. `predict` and `margins` are available on the core estimator layer only.
+- **Full command surfaces for community commands** — `reghdfe`, `ivreghdfe`, `ppmlhdfe`, `did_imputation`, `eventstudyinteract`, `csdid`, and `rdrobust` are implemented as **verified high-frequency subsets**, not complete Stata command reproductions. Unsupported options are explicitly rejected rather than silently ignored.
+
+### Completeness legend
+
+- **Stable** — synthetic + real-data dual-run verified; core API is unlikely to change.
+- **Alpha** — high-frequency paths are implemented and verified, but the command surface is still a subset of the full Stata community command.
+- **Alpha — Partial** — a verifiable implementation exists, but large functional areas are still missing (e.g., fuzzy RD for `rdrobust`, multi-way clustering).
+
+See the [Command Support Matrix](./docs/command-support-matrix/README.md) for the per-command detailed status.
+
+---
+
+## Installation
+
+```bash
+pip install StataFlow
+```
+
+Requirements: Python 3.10+, NumPy, pandas, SciPy.
+
+For development (editable install from source):
+
+```bash
+git clone https://github.com/ZhenHaoFu810/StataFlow.git
+cd StataFlow
+pip install -e .
+```
+
+---
+
+## Quick start
+
+### Stata-compatible command layer (recommended)
+
+All `compat.stata` wrappers return a `ResultSchema` object with coefficients, standard errors, and fit statistics. They do **not** expose `.predict()` or `.margins()` directly—use the core estimator layer below for post-estimation.
+
+```python
+import pandas as pd
+from stataflow.compat.stata import regress, reghdfe, ivregress_2sls, logit
+
+# OLS with robust standard errors
+result = regress(df, y="wage", x=["edu", "exper"], vce="robust")
+
+# High-dimensional fixed effects (reghdfe)
+result = reghdfe(
+    df, y="wage", x=["edu", "exper"],
+    absorb="firm_id year_id", vce="cluster", cluster="industry"
+)
+
+# Factor-variable syntax in HDFE
+result = reghdfe(
+    df, y="wage", x=["i.industry##c.post"], absorb="firm_id year_id"
+)
+
+# 2SLS
+result = ivregress_2sls(
+    df, y="lwage", x_exog=["edu"], x_endog=["exper"],
+    instruments=["age", "kidslt6"], vce="robust"
+)
+
+# Logit
+result = logit(df, y="inlf", x=["nwifeinc", "educ", "exper"])
+```
+
+For runnable examples, see the [`examples/`](./examples/) directory:
+- [`examples/demo_regress.py`](./examples/demo_regress.py)
+- [`examples/demo_reghdfe.py`](./examples/demo_reghdfe.py)
+- [`examples/demo_ppmlhdfe.py`](./examples/demo_ppmlhdfe.py)
+- [`examples/demo_ivregress_2sls.py`](./examples/demo_ivregress_2sls.py)
+
+### Native Python estimator layer (advanced)
+
+```python
+from stataflow import OLS, FixedEffectsOLS, AbsorbingOLS, Logit, IV2SLS
+
+model = OLS(data=df, y="wage", x=["edu", "exper"])
+result = model.fit(vce="robust")
+```
+
+---
+
+## Supported commands
+
+| Command | Python entry | Core capabilities |
+|---------|--------------|-------------------|
+| `regress` | `stataflow.compat.stata.regress` | OLS, robust, cluster, aweight |
+| `xtreg, fe` | `stataflow.compat.stata.xtreg_fe` | Fixed effects (within), cluster |
+| `areg` | `stataflow.compat.stata.areg` | Single absorb variable FE |
+| `reghdfe` | `stataflow.compat.stata.reghdfe` | 1-2 group HDFE, cluster, singleton drop |
+| `ivregress 2sls` | `stataflow.compat.stata.ivregress_2sls` | 2SLS, robust, cluster |
+| `ivreghdfe` | `stataflow.compat.stata.ivreghdfe` | IV + 1-2 group HDFE, cluster |
+| `logit` | `stataflow.compat.stata.logit` | MLE, robust, cluster |
+| `probit` | `stataflow.compat.stata.probit` | MLE, robust, cluster |
+| `poisson` | `stataflow.compat.stata.poisson` | MLE, robust, cluster |
+| `ppmlhdfe` | `stataflow.compat.stata.ppmlhdfe` | PPML + 1-2 group HDFE |
+| `did_imputation` | `stataflow.compat.stata.did_imputation` | BJS DID imputation |
+| `eventstudyinteract` | `stataflow.compat.stata.eventstudyinteract` | Sun & Abraham IW estimator |
+| `csdid` | `stataflow.compat.stata.csdid` | Callaway-Sant'Anna DID (`method="reg"` only) |
+| `rdrobust` | `stataflow.compat.stata.rdrobust` | Sharp RD local polynomial (`bwselect="mserd"`, `covs`) |
+
+Full details: [`docs/command-support-matrix/README.md`](./docs/command-support-matrix/README.md)
+
+---
+
+## Validation philosophy
+
+Every public command is validated with **two lines of evidence**:
+
+1. **Synthetic / controlled cases** — formula, degrees of freedom, sample screening, edge cases.
+2. **Real public datasets** — field-level comparison against Stata 17 on openly available economic/financial data.
+
+A command is considered "done" only when both lines pass and the source-to-Python mapping is documented. We do not accept "statistical equivalence" without explicit mathematical or source-code justification.
+
+Public evidence and results are available in `research/results/validation/`.
+
+### Running tests
+
+```bash
+# Unit and integration tests (fast)
+pytest tests/ -v --ignore=tests/golden/
+
+# Golden dual-run tests (require Stata 17)
+pytest tests/golden/ -v
+```
+
+---
+
+## Project structure
+
+- **`src/stataflow/estimators/`** — Core Python estimators (`OLS`, `AbsorbingOLS`, `Logit`, `PPMLHDFE`, `DIDImputation`, etc.)
+- **`src/stataflow/compat/stata/`** — Stata command wrappers (`regress()`, `reghdfe()`, `ivregress_2sls()`, etc.)
+- **`docs/command-support-matrix/`** — Per-command support matrices
+- **`examples/`** — Runnable demonstration scripts
+- **`tests/`** — Unit and integration tests
+
+---
+
+## Default target version
+
+**Stata 17**
+
+---
+
+## Documentation
+
+- [User guide](./docs/USER_GUIDE.md)
+- [Chinese user guide](./docs/USER_GUIDE.zh-CN.md)
+- [Cookbook](./docs/cookbook.md)
+- [Chinese cookbook](./docs/cookbook.zh-CN.md)
+- [Command support matrices](./docs/command-support-matrix/README.md)
+
+---
+
+## Governance
+
+- **Codex** — project goals, architecture, review gates, and statistical-dispute arbitration.
+- **Claude Code** — implementation, testing, and evidence backfill.

{stataflow-0.1.2 → stataflow-0.1.4}/pyproject.toml

@@ -4,15 +4,16 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "StataFlow"
-version = "0.1.2"
-description = "Stata2Python: A Python econometrics toolkit aligned with Stata 17"
+version = "0.1.4"
+description = "StataFlow: A Python econometrics toolkit aligned with Stata 17"
 readme = "README.md"
-license = "MIT"
+license = {text = "MIT"}
 authors = [{name = "Zhenhao Fu", email = "zhenhaofu2001@gmail.com"}]
 keywords = ["econometrics", "stata", "regression", "fixed-effects", "panel-data"]
 classifiers = [
     "Development Status :: 3 - Alpha",
     "Intended Audience :: Science/Research",
+    "License :: OSI Approved :: MIT License",
     "Programming Language :: Python :: 3",
     "Programming Language :: Python :: 3.10",
     "Programming Language :: Python :: 3.11",

stataflow-0.1.4/src/StataFlow.egg-info/PKG-INFO

@@ -0,0 +1,202 @@
+Metadata-Version: 2.4
+Name: StataFlow
+Version: 0.1.4
+Summary: StataFlow: A Python econometrics toolkit aligned with Stata 17
+Author-email: Zhenhao Fu <zhenhaofu2001@gmail.com>
+License: MIT
+Project-URL: Homepage, https://github.com/ZhenHaoFu810/StataFlow
+Project-URL: Repository, https://github.com/ZhenHaoFu810/StataFlow
+Project-URL: Issues, https://github.com/ZhenHaoFu810/StataFlow/issues
+Keywords: econometrics,stata,regression,fixed-effects,panel-data
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Scientific/Engineering :: Mathematics
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: numpy>=1.24
+Requires-Dist: pandas>=2.0
+Requires-Dist: scipy>=1.10
+Requires-Dist: pyyaml>=6.0
+Provides-Extra: dev
+Requires-Dist: pytest>=7.0; extra == "dev"
+Requires-Dist: pytest-cov>=4.0; extra == "dev"
+Dynamic: license-file
+
+# StataFlow
+
+StataFlow (`stataflow`) is a Python econometrics toolkit that reproduces Stata 17 estimation results with high precision. It provides both a **Stata-compatible command layer** (for researchers migrating from Stata) and a **native Python estimator layer** (for advanced users who want direct control).
+
+## What you can do today
+
+- Run Stata-style commands in Python: `regress`, `reghdfe`, `ivregress 2sls`, `logit`, `ppmlhdfe`, `did_imputation`, `csdid`, `rdrobust`, and more.
+- Obtain coefficients, standard errors, t/z-statistics, p-values, and confidence intervals that are field-level verified against Stata 17.
+- Work with high-dimensional fixed effects (HDFE), IV/2SLS, binary/count models, and DID/event-study estimators.
+- Use Stata-style factor-variable syntax (`i.group##c.post`, `c.x1#c.x2`, `x1##x2`) and space-separated absorb strings directly in wrapper commands. Bare variables inside `#` / `##` are treated as continuous, matching common Stata usage.
+
+## What is not yet supported
+
+- **Multi-way clustering** — only single-cluster robust inference is available.
+- **Direct post-estimation on wrapper returns** — the `compat.stata` wrappers return `ResultSchema` result objects. `predict` and `margins` are available on the core estimator layer only.
+- **Full command surfaces for community commands** — `reghdfe`, `ivreghdfe`, `ppmlhdfe`, `did_imputation`, `eventstudyinteract`, `csdid`, and `rdrobust` are implemented as **verified high-frequency subsets**, not complete Stata command reproductions. Unsupported options are explicitly rejected rather than silently ignored.
+
+### Completeness legend
+
+- **Stable** — synthetic + real-data dual-run verified; core API is unlikely to change.
+- **Alpha** — high-frequency paths are implemented and verified, but the command surface is still a subset of the full Stata community command.
+- **Alpha — Partial** — a verifiable implementation exists, but large functional areas are still missing (e.g., fuzzy RD for `rdrobust`, multi-way clustering).
+
+See the [Command Support Matrix](./docs/command-support-matrix/README.md) for the per-command detailed status.
+
+---
+
+## Installation
+
+```bash
+pip install StataFlow
+```
+
+Requirements: Python 3.10+, NumPy, pandas, SciPy.
+
+For development (editable install from source):
+
+```bash
+git clone https://github.com/ZhenHaoFu810/StataFlow.git
+cd StataFlow
+pip install -e .
+```
+
+---
+
+## Quick start
+
+### Stata-compatible command layer (recommended)
+
+All `compat.stata` wrappers return a `ResultSchema` object with coefficients, standard errors, and fit statistics. They do **not** expose `.predict()` or `.margins()` directly—use the core estimator layer below for post-estimation.
+
+```python
+import pandas as pd
+from stataflow.compat.stata import regress, reghdfe, ivregress_2sls, logit
+
+# OLS with robust standard errors
+result = regress(df, y="wage", x=["edu", "exper"], vce="robust")
+
+# High-dimensional fixed effects (reghdfe)
+result = reghdfe(
+    df, y="wage", x=["edu", "exper"],
+    absorb="firm_id year_id", vce="cluster", cluster="industry"
+)
+
+# Factor-variable syntax in HDFE
+result = reghdfe(
+    df, y="wage", x=["i.industry##c.post"], absorb="firm_id year_id"
+)
+
+# 2SLS
+result = ivregress_2sls(
+    df, y="lwage", x_exog=["edu"], x_endog=["exper"],
+    instruments=["age", "kidslt6"], vce="robust"
+)
+
+# Logit
+result = logit(df, y="inlf", x=["nwifeinc", "educ", "exper"])
+```
+
+For runnable examples, see the [`examples/`](./examples/) directory:
+- [`examples/demo_regress.py`](./examples/demo_regress.py)
+- [`examples/demo_reghdfe.py`](./examples/demo_reghdfe.py)
+- [`examples/demo_ppmlhdfe.py`](./examples/demo_ppmlhdfe.py)
+- [`examples/demo_ivregress_2sls.py`](./examples/demo_ivregress_2sls.py)
+
+### Native Python estimator layer (advanced)
+
+```python
+from stataflow import OLS, FixedEffectsOLS, AbsorbingOLS, Logit, IV2SLS
+
+model = OLS(data=df, y="wage", x=["edu", "exper"])
+result = model.fit(vce="robust")
+```
+
+---
+
+## Supported commands
+
+| Command | Python entry | Core capabilities |
+|---------|--------------|-------------------|
+| `regress` | `stataflow.compat.stata.regress` | OLS, robust, cluster, aweight |
+| `xtreg, fe` | `stataflow.compat.stata.xtreg_fe` | Fixed effects (within), cluster |
+| `areg` | `stataflow.compat.stata.areg` | Single absorb variable FE |
+| `reghdfe` | `stataflow.compat.stata.reghdfe` | 1-2 group HDFE, cluster, singleton drop |
+| `ivregress 2sls` | `stataflow.compat.stata.ivregress_2sls` | 2SLS, robust, cluster |
+| `ivreghdfe` | `stataflow.compat.stata.ivreghdfe` | IV + 1-2 group HDFE, cluster |
+| `logit` | `stataflow.compat.stata.logit` | MLE, robust, cluster |
+| `probit` | `stataflow.compat.stata.probit` | MLE, robust, cluster |
+| `poisson` | `stataflow.compat.stata.poisson` | MLE, robust, cluster |
+| `ppmlhdfe` | `stataflow.compat.stata.ppmlhdfe` | PPML + 1-2 group HDFE |
+| `did_imputation` | `stataflow.compat.stata.did_imputation` | BJS DID imputation |
+| `eventstudyinteract` | `stataflow.compat.stata.eventstudyinteract` | Sun & Abraham IW estimator |
+| `csdid` | `stataflow.compat.stata.csdid` | Callaway-Sant'Anna DID (`method="reg"` only) |
+| `rdrobust` | `stataflow.compat.stata.rdrobust` | Sharp RD local polynomial (`bwselect="mserd"`, `covs`) |
+
+Full details: [`docs/command-support-matrix/README.md`](./docs/command-support-matrix/README.md)
+
+---
+
+## Validation philosophy
+
+Every public command is validated with **two lines of evidence**:
+
+1. **Synthetic / controlled cases** — formula, degrees of freedom, sample screening, edge cases.
+2. **Real public datasets** — field-level comparison against Stata 17 on openly available economic/financial data.
+
+A command is considered "done" only when both lines pass and the source-to-Python mapping is documented. We do not accept "statistical equivalence" without explicit mathematical or source-code justification.
+
+Public evidence and results are available in `research/results/validation/`.
+
+### Running tests
+
+```bash
+# Unit and integration tests (fast)
+pytest tests/ -v --ignore=tests/golden/
+
+# Golden dual-run tests (require Stata 17)
+pytest tests/golden/ -v
+```
+
+---
+
+## Project structure
+
+- **`src/stataflow/estimators/`** — Core Python estimators (`OLS`, `AbsorbingOLS`, `Logit`, `PPMLHDFE`, `DIDImputation`, etc.)
+- **`src/stataflow/compat/stata/`** — Stata command wrappers (`regress()`, `reghdfe()`, `ivregress_2sls()`, etc.)
+- **`docs/command-support-matrix/`** — Per-command support matrices
+- **`examples/`** — Runnable demonstration scripts
+- **`tests/`** — Unit and integration tests
+
+---
+
+## Default target version
+
+**Stata 17**
+
+---
+
+## Documentation
+
+- [User guide](./docs/USER_GUIDE.md)
+- [Chinese user guide](./docs/USER_GUIDE.zh-CN.md)
+- [Cookbook](./docs/cookbook.md)
+- [Chinese cookbook](./docs/cookbook.zh-CN.md)
+- [Command support matrices](./docs/command-support-matrix/README.md)
+
+---
+
+## Governance
+
+- **Codex** — project goals, architecture, review gates, and statistical-dispute arbitration.
+- **Claude Code** — implementation, testing, and evidence backfill.

stataflow-0.1.4/src/StataFlow.egg-info/SOURCES.txt

@@ -0,0 +1,47 @@
+LICENSE
+README.md
+pyproject.toml
+src/StataFlow.egg-info/PKG-INFO
+src/StataFlow.egg-info/SOURCES.txt
+src/StataFlow.egg-info/dependency_links.txt
+src/StataFlow.egg-info/requires.txt
+src/StataFlow.egg-info/top_level.txt
+src/stataflow/__init__.py
+src/stataflow/postestimation.py
+src/stataflow/compat/__init__.py
+src/stataflow/compat/stata/__init__.py
+src/stataflow/compat/stata/did.py
+src/stataflow/compat/stata/factor_variables.py
+src/stataflow/compat/stata/glm.py
+src/stataflow/compat/stata/hdfe.py
+src/stataflow/compat/stata/iv.py
+src/stataflow/compat/stata/linear.py
+src/stataflow/compat/stata/rdrobust.py
+src/stataflow/estimators/__init__.py
+src/stataflow/estimators/absorbing_ols.py
+src/stataflow/estimators/csdid.py
+src/stataflow/estimators/did_imputation.py
+src/stataflow/estimators/eventstudyinteract.py
+src/stataflow/estimators/fe.py
+src/stataflow/estimators/glm.py
+src/stataflow/estimators/iv.py
+src/stataflow/estimators/ols.py
+src/stataflow/estimators/ppmlhdfe.py
+src/stataflow/estimators/rdrobust.py
+src/stataflow/results/__init__.py
+src/stataflow/results/result.py
+src/stataflow/stata_runner/__init__.py
+src/stataflow/stata_runner/runner.py
+tests/test_compat_stata_did.py
+tests/test_compat_stata_glm.py
+tests/test_compat_stata_hdfe.py
+tests/test_compat_stata_iv.py
+tests/test_compat_stata_linear.py
+tests/test_factor_variables.py
+tests/test_hdfe_synthetic.py
+tests/test_ols_against_statsmodels.py
+tests/test_rdrobust.py
+tests/test_result_schema.py
+tests/test_smoke.py
+tests/test_stata_runner.py
+tests/test_validation_summary.py

stataflow-0.1.4/src/StataFlow.egg-info/top_level.txt

@@ -0,0 +1 @@
+stataflow