release: v1.0.0

Author: YuminosukeSato

.github/workflows/release.yml

@@ -54,8 +54,63 @@ jobs:
           name: sdist
           path: dist/*.tar.gz
 
+  smoke-test-wheels:
+    needs: build-wheels
+    runs-on: ${{ matrix.os }}
+    strategy:
+      fail-fast: false
+      matrix:
+        include:
+          - os: ubuntu-latest
+            target: x86_64
+          - os: macos-14
+            target: aarch64
+          - os: macos-latest
+            target: x86_64
+          - os: windows-latest
+            target: x64
+    steps:
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+      - uses: actions/download-artifact@v4
+        with:
+          name: wheels-${{ matrix.os }}-${{ matrix.target }}
+          path: dist
+      - name: Install built wheel
+        shell: python
+        run: |
+          import glob
+          import subprocess
+          import sys
+
+          wheels = glob.glob("dist/*.whl")
+          assert wheels, "No wheel artifact downloaded"
+          subprocess.check_call([sys.executable, "-m", "pip", "install", wheels[0]])
+      - name: Smoke test wheel
+        shell: python
+        run: |
+          import numpy as np
+          import pandas as pd
+          from causal_impact import CausalImpact, ModelOptions
+
+          dates = pd.date_range("2020-01-01", periods=24, freq="D")
+          x = np.linspace(0.0, 1.0, 24)
+          y = 5.0 + 0.2 * np.arange(24) + 1.5 * x
+          y[18:] += 2.0
+          data = pd.DataFrame({"y": y, "x1": x}, index=dates)
+
+          ci = CausalImpact(
+              data,
+              ["2020-01-01", "2020-01-18"],
+              ["2020-01-19", "2020-01-24"],
+              model_args=ModelOptions(niter=100, nwarmup=50, seed=42),
+          )
+          assert ci.summary()
+          assert len(ci.inferences.columns) >= 11
+
   publish:
-    needs: [build-wheels, build-sdist]
+    needs: [build-sdist, smoke-test-wheels]
     runs-on: ubuntu-latest
     environment: pypi
     permissions:

.gitignore

@@ -4,6 +4,7 @@ __pycache__/
 *.pyc
 *.pyo
 *.so
+*.dSYM/
 *.egg-info/
 dist/
 build/

CHANGELOG.md

@@ -0,0 +1,31 @@
+# Changelog
+
+All notable changes to this project will be documented in this file.
+
+The format is based on Keep a Changelog.
+
+## [1.0.0] - 2026-03-23
+
+### Added
+
+- Added `state_model` with `local_level` and `local_linear_trend` support.
+- Added `actual` and `predictions_sd` to `ci.inferences`.
+- Added `py.typed` for PEP 561 type discovery.
+- Added wheel smoke tests to the release workflow.
+
+### Changed
+
+- Made summary and report CI labels respect `alpha`.
+- Unified `expected_model_size` defaults at `2`.
+- Clarified installation guidance around binary wheels versus source builds.
+
+## [0.3.0]
+
+### Added
+
+- Added dynamic regression with time-varying coefficients.
+- Added seasonal components and MkDocs documentation site.
+
+### Changed
+
+- Improved numerical equivalence coverage against R fixtures.

Cargo.lock

@@ -1,6 +1,6 @@
 [package]
 name = "causal_impact_core"
-version = "0.3.0"
+version = "1.0.0"
 edition = "2021"
 
 [lib]

Cargo.toml

@@ -17,7 +17,9 @@ If any of these assumptions are violated, the causal estimate will be unreliable
 
 ## Installation
 
-Requires Python 3.10+ and a Rust toolchain.
+Requires Python 3.10+.
+Binary wheels are intended for supported platforms, so Rust is only required
+when building from source.
 
 ```bash
 pip install bsts-causalimpact
@@ -113,12 +115,12 @@ Tests run on every commit with seed-fixed MCMC for deterministic reproduction.
 | Post-period Random Walk propagation | Matching | Forward simulation from last pre-period state |
 | Data standardization (standardize.data=TRUE) | Matching | (y - mean) / sd using pre-period moments |
 | prior.level.sd = 0.01 | Matching | Same default, same semantics |
-| Spike-and-slab variable selection | Partial | Coordinate-wise sampling works; prior parameters (expected.r2, prior.df) are approximate |
-| expected.model.size | Partial | `CausalImpact` preserves the legacy default `2`; `ModelOptions` keeps explicit default `1` |
-| expected.r2 = 0.8, prior.df = 50 | Partial | Static regression prior is tuned for close R parity, not a byte-for-byte port |
+| Spike-and-slab variable selection | Matching | Coordinate-wise sampling with StudentSpikeSlabPrior defaults (`expected.r2=0.8`, `prior.df=50`, `prior.information.weight=0.01`, `diagonal.shrinkage=0.5`) |
+| expected.model.size | Matching | Unified default `2` in `CausalImpact` and `ModelOptions` |
+| expected.r2 = 0.8, prior.df = 50 | Matching | Same documented residual variance prior defaults as BoomSpikeSlab / bsts |
 | Seasonal component (`nseasons`, `season_duration`) | Supported | R-compatible API with seasonal fixture coverage |
 | Dynamic regression | Supported | Time-varying coefficients via random-walk FFBS; `dynamic_regression=True` |
-| Local linear trend | Planned | R uses AddLocalLevel only by default; trend option exists but not ported |
+| Local linear trend | Supported | Opt in with `state_model="local_linear_trend"` |
 
 Covariate CI bounds are enforced twice: the legacy parity fixture remains tighter than
 Phase 2 requirements, and a separate Phase 2 acceptance test keeps the threshold at `±3%`.
@@ -145,10 +147,11 @@ Phase 2 requirements, and a separate Phase 2 acceptance test keeps the threshold
 | `seed` | 0 | Random seed for reproducibility |
 | `prior_level_sd` | 0.01 | Prior standard deviation for the local level |
 | `standardize_data` | `True` | Standardize data before fitting |
-| `expected_model_size` | 2 | Expected number of active covariates (spike-and-slab prior); `ModelOptions` keeps `1` |
+| `expected_model_size` | 2 | Expected number of active covariates (spike-and-slab prior) |
 | `nseasons` | `None` | Optional seasonal cycle count (R-compatible API) |
 | `season_duration` | `None` | Optional duration of each seasonal block; defaults to `1` when `nseasons` is set |
 | `dynamic_regression` | `False` | Enable time-varying regression coefficients (random-walk beta) |
+| `state_model` | `"local_level"` | `"local_level"` or `"local_linear_trend"` |
 
 #### Methods and Properties
 
@@ -157,7 +160,7 @@ Phase 2 requirements, and a separate Phase 2 acceptance test keeps the threshold
 | `summary(output="summary")` | `str` | Tabular summary of causal effects |
 | `report()` | `str` | Narrative interpretation of results |
 | `plot(metrics=None)` | `Figure` | Matplotlib figure with original/pointwise/cumulative panels |
-| `inferences` | `DataFrame` | Per-timestep effects, predictions, and credible intervals |
+| `inferences` | `DataFrame` | Per-timestep actuals, predictions, prediction s.d., and effect intervals |
 | `summary_stats` | `dict` | Aggregate statistics (effect mean, CI, p-value, etc.) |
 | `posterior_inclusion_probs` | `ndarray \| None` | Posterior inclusion probability per covariate |
 

README.md

@@ -30,7 +30,7 @@ ci = CausalImpact(data, pre_period, post_period, model_args=None, alpha=0.05)
 
 | Property | Type | Description |
 |---|---|---|
-| `inferences` | `DataFrame` | Per-timestep effects, predictions, and credible intervals |
+| `inferences` | `DataFrame` | Per-timestep actuals, predictions, prediction s.d., and effect intervals |
 | `summary_stats` | `dict` | Aggregate statistics (effect mean, CI, p-value, etc.) |
 | `posterior_inclusion_probs` | `ndarray \| None` | Posterior inclusion probability per covariate (requires covariates) |
 
@@ -53,15 +53,12 @@ ci = CausalImpact(data, pre_period, post_period, model_args=opts)
 | `seed` | `int` | 0 | Random seed for reproducibility |
 | `prior_level_sd` | `float` | 0.01 | Prior standard deviation for the local level |
 | `standardize_data` | `bool` | `True` | Standardize data before fitting |
-| `expected_model_size` | `int` | 1 | Expected number of active covariates for spike-and-slab prior |
+| `expected_model_size` | `int` | 2 | Expected number of active covariates for spike-and-slab prior |
+| `dynamic_regression` | `bool` | `False` | Enable time-varying regression coefficients |
+| `state_model` | `str` | `"local_level"` | `"local_level"` or `"local_linear_trend"` |
 | `nseasons` | `int \| None` | `None` | Seasonal cycle count |
 | `season_duration` | `int \| None` | `None` | Duration of each seasonal block; defaults to 1 when `nseasons` is set |
 
-!!! note "expected_model_size defaults"
-    `CausalImpact` sets `expected_model_size=2` by default (matching R).
-    `ModelOptions` keeps `expected_model_size=1` as the explicit default.
-    When passing a `ModelOptions` instance, the `ModelOptions` value takes precedence.
-
 ## `CausalImpactResults`
 
 Returned by `ci._results`. A frozen dataclass containing all computed quantities.
@@ -72,14 +69,15 @@ Returned by `ci._results`. A frozen dataclass containing all computed quantities
 |---|---|---|
 | `actual` | `ndarray` | Observed y values in the post period |
 | `point_effects` | `ndarray` | Mean effect per time point |
-| `point_effect_lower` | `ndarray` | Lower 95% CI per time point |
-| `point_effect_upper` | `ndarray` | Upper 95% CI per time point |
+| `point_effect_lower` | `ndarray` | Lower pointwise credible interval per time point |
+| `point_effect_upper` | `ndarray` | Upper pointwise credible interval per time point |
 | `point_effect_mean` | `float` | Mean of point effects across time |
 | `ci_lower` | `float` | Lower CI bound on average effect |
 | `ci_upper` | `float` | Upper CI bound on average effect |
 | `cumulative_effect_total` | `float` | Total cumulative effect |
 | `relative_effect_mean` | `float` | Relative effect (effect / predicted) |
 | `p_value` | `float` | Bayesian one-sided tail probability |
 | `predictions_mean` | `ndarray` | Mean counterfactual prediction |
+| `predictions_sd` | `ndarray` | Posterior standard deviation of the counterfactual prediction |
 | `predictions_lower` | `ndarray` | Lower CI on counterfactual |
 | `predictions_upper` | `ndarray` | Upper CI on counterfactual |

docs/api.md

@@ -7,9 +7,9 @@ Comparison of features between R CausalImpact (bsts 1.4.1) and this Python imple
 | Feature | R | Python | Notes |
 |---|---|---|---|
 | Local level | Yes | Yes | Identical algorithm |
-| Local linear trend | Yes | No | Not yet implemented |
+| Local linear trend | Yes | Yes | `state_model="local_linear_trend"` |
 | Seasonality | Yes | Yes | R-compatible API with seasonal fixture coverage |
-| Dynamic regression | Yes | No | Not yet implemented |
+| Dynamic regression | Yes | Yes | `dynamic_regression=True` |
 | Regression (static) | Yes | Yes | Identical algorithm |
 
 ## MCMC Parameters
@@ -21,7 +21,8 @@ Comparison of features between R CausalImpact (bsts 1.4.1) and this Python imple
 | season.duration | Yes | Yes | `ModelOptions.season_duration` or `model_args["season.duration"]` |
 | prior.level.sd | Yes | Yes | Same default (0.01) |
 | standardize.data | Yes | Yes | Same default (True) |
-| expected.model.size | Yes | Yes | Legacy `CausalImpact` default is 2; `ModelOptions` keeps 1 |
+| expected.model.size | Yes | Yes | Unified default `2` |
+| state model selection | Via bsts state spec | Yes | `state_model="local_level"` or `"local_linear_trend"` |
 
 ## Warmup Semantics
 

docs/compatibility-matrix.md

@@ -11,7 +11,8 @@ No TensorFlow required. 10-30x faster than R.
 pip install bsts-causalimpact
 ```
 
-Requires Python 3.10+ and a Rust toolchain (only for building from source).
+Requires Python 3.10+.
+Rust is only needed when building from source instead of installing a wheel.
 
 ## Example: Measuring the Effect of an Intervention
 
@@ -168,6 +169,8 @@ print(ci.posterior_inclusion_probs)
 | `prior_level_sd` | 0.01 | Prior standard deviation for the local level |
 | `standardize_data` | `True` | Standardize data before fitting |
 | `expected_model_size` | 2 | Expected number of active covariates (spike-and-slab prior) |
+| `dynamic_regression` | `False` | Enable time-varying regression coefficients |
+| `state_model` | `"local_level"` | `"local_level"` or `"local_linear_trend"` |
 | `nseasons` | `None` | Seasonal cycle count |
 | `season_duration` | `None` | Duration of each seasonal block (defaults to 1 when `nseasons` is set) |
 

docs/index.md

@@ -43,7 +43,7 @@ fig = ci.plot()
 | `season.duration` | `season_duration` or `model_args["season.duration"]` | `1` when `nseasons` is set |
 | `prior.level.sd` | `prior_level_sd` | 0.01 |
 | `standardize.data` | `standardize_data` | True |
-| `expected.model.size` | `expected_model_size` | 2 in `CausalImpact`; `ModelOptions` keeps 1 |
+| `expected.model.size` | `expected_model_size` | 2 |
 
 ## Data Format
 
@@ -85,9 +85,11 @@ This library verifies ±3% agreement with R CausalImpact on point estimates and
 
 Differences arise from independent RNG implementations (R's `set.seed` vs Rust's `ChaCha8Rng`), not from algorithmic differences.
 
+Use `model_args={"state_model": "local_linear_trend"}` when you want the
+optional local linear trend state instead of the default local level model.
+
 ## What Is Not Supported
 
 - Custom bsts model objects
-- `model.args$dynamic.regression`
 
 These features may be added in future versions.