feat: v0.9.15 — BayesianMTEResult.tidy(terms=[...]) multi-row output

Completes the broom-pipeline integration of v0.9.13's per-population
ATT/ATU uncertainty. Users can now pd.concat ATE/ATT/ATU rows across
fits in a single call.

Added:
- BayesianMTEResult.tidy(conf_level=None, terms=None) override.
  terms=None (default) unchanged → single ATE row. Single-string
  terms='ate'/'att'/'atu' → single row of that term. List like
  ['ate', 'att', 'atu'] → multi-row DataFrame with 9-column schema.
  Invalid term names raise ValueError.

- BayesianMTEResult now carries att_prob_positive and
  atu_prob_positive fields (NaN-defaulted for pre-v0.9.15
  snapshot back-compat). Populated by _integrated_effect from the
  per-draw ATT/ATU posteriors.

- _integrated_effect returns a 5-tuple (mean, sd, hdi_lower,
  hdi_upper, prob_positive) instead of 4-tuple. Caller updated
  accordingly.

Round B review: 1 HIGH. Round C fixed:
- HIGH-1: default tidy() emits term='ate (integrated mte)' (via
  parent estimand.lower()) but tidy(terms='ate') emitted short
  literal 'ate'. Mixing both call styles in a concat pipeline
  would have produced inconsistent labels for the same term.
  Fixed by routing the 'ate' branch through
  self.estimand.lower() so both paths produce byte-identical
  rows. ATT / ATU keep short labels (no parent-default precedent;
  short is natural for new terms). Regression test
  test_tidy_ate_matches_default_path asserts exact parity.

- Round C re-review: zero ship-blockers.

Tests:
- tests/test_bayes_mte_tidy.py (13 tests) — back-compat default,
  single-term paths, multi-row preservation of order, concat
  workflow, invalid-term rejection (single + mixed with valid),
  NaN prob_positive stub back-compat, prob_positive scalars
  populated on real fits, default-vs-explicit label byte-parity
  (Round-C regression).
- Bayesian family suite: 101/101 focused tests green in 2:14.

Design spec: docs/superpowers/specs/2026-04-20-v0915-tidy-multiterm.md

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Author: brycewang-stanford

CHANGELOG.md

@@ -2,6 +2,64 @@
 
 All notable changes to StatsPAI will be documented in this file.
 
+## [0.9.15] - 2026-04-20 — Multi-term `tidy(terms=[...])` + ATT/ATU prob_positive
+
+Completes the broom-pipeline integration of v0.9.13's per-population
+ATT/ATU uncertainty. Users can now `pd.concat` ATE/ATT/ATU rows
+across fits in one call.
+
+### Added (0.9.15)
+
+- **`BayesianMTEResult.tidy(conf_level=None, terms=None)`** override:
+  - `terms=None` (default) — unchanged, single ATE row.
+  - `terms='ate' | 'att' | 'atu'` — single row of that term.
+  - `terms=['ate', 'att', 'atu']` — multi-row DataFrame.
+  - Invalid names → clear `ValueError`.
+
+- **Two new result fields**: `att_prob_positive`, `atu_prob_positive`
+  (NaN-defaulted for pre-v0.9.15 snapshot compatibility). Populated
+  by `_integrated_effect` from per-draw ATT/ATU posteriors.
+
+- **`_integrated_effect` returns 5-tuple** `(mean, sd, hdi_lower,
+  hdi_upper, prob_positive)`. Caller unpacks + passes to the result.
+
+### Round-B review found 1 HIGH; Round-C fixed
+
+- **HIGH-1** — label divergence: default `tidy()` emits
+  `term='ate (integrated mte)'` (via parent `estimand.lower()`),
+  but `tidy(terms='ate')` emitted the short literal `'ate'`. Byte-
+  compat broken when a user mixed both call styles inside
+  `pd.concat`. **Fixed** — `_row('ate')` now also uses
+  `self.estimand.lower()` so both paths produce identical rows.
+  ATT / ATU rows keep their short labels (no parent-default
+  precedent; short is the natural broom shape for new terms).
+
+- Round C: 0 blockers.
+
+### Tests (0.9.15)
+
+- `tests/test_bayes_mte_tidy.py` (13 tests) — back-compat default
+  schema, single-term paths for all three labels, multi-row order
+  preservation, concat workflow, invalid-term + mixed-valid
+  rejection, NaN prob_positive stub back-compat, prob_positive
+  scalars populated on real fits, **default-vs-explicit label
+  byte-parity** (Round-C regression).
+- Bayesian family suite: 101/101 focused tests green.
+
+### Design spec (0.9.15)
+
+- `docs/superpowers/specs/2026-04-20-v0915-tidy-multiterm.md`
+
+### Non-goals (0.9.15)
+
+- Multi-term `.tidy()` on other Bayesian estimators — DID/RD/IV
+  have no ATT/ATU concept; the primary-estimand row is already
+  what they emit.
+- Full bivariate-normal HV model.
+- Rust Phase 2.
+
+---
+
 ## [0.9.14] - 2026-04-20 — Summary rendering completes v0.9.13 spec §3.3
 
 Tiny patch release. Completes the "ATT/ATU in `summary()`" promise

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "StatsPAI"
-version = "0.9.14"
+version = "0.9.15"
 description = "The Agent-Native Causal Inference & Econometrics Toolkit for Python"
 readme = "README.md"
 license = {text = "MIT"}

src/statspai/__init__.py

@@ -22,7 +22,7 @@
 >>> sp.outreg2(result, filename="results.xlsx")
 """
 
-__version__ = "0.9.14"
+__version__ = "0.9.15"
 __author__ = "Biaoyue Wang"
 __email__ = "brycew6m@stanford.edu"
 
@@ -33,6 +33,8 @@
 from .causal.forest_inference import (
     calibration_test, test_calibration, rate, honest_variance,
 )
+from .causal.multi_arm_forest import multi_arm_forest, MultiArmForestResult
+from .causal.iv_forest import iv_forest, IVForestResult
 from .did import (
     did, did_2x2, ddd, callaway_santanna, sun_abraham,
     bacon_decomposition, honest_did, breakdown_m, event_study,
@@ -78,6 +80,7 @@
     optimal_match, cardinality_match,
     OptimalMatchResult, CardinalityMatchResult,
     overlap_weights, cbps,
+    genmatch, GenMatchResult,
 )
 from .dml import dml, DoubleML, DoubleMLPLR, DoubleMLIRM, DoubleMLPLIV, DoubleMLIIVM
 from .deepiv import deepiv, DeepIV
@@ -110,6 +113,7 @@
     proximal, ProximalCausalInference,
     negative_control_outcome, negative_control_exposure,
     double_negative_control, NegativeControlResult,
+    proximal_regression, ProximalRegResult,
 )
 from .principal_strat import (
     principal_strat, PrincipalStratResult, survivor_average_causal_effect,
@@ -124,6 +128,7 @@
     moran, moran_local, geary, getis_ord_g, getis_ord_local, join_counts,
     moran_plot, lisa_cluster_map,
     lm_tests, moran_residuals, impacts,
+    spatial_did, SpatialDiDResult, spatial_iv, SpatialIVResult,
 )
 from . import spatial
 # NOTE: `from . import iv` would be shadowed by the `iv` function imported
@@ -161,20 +166,41 @@
 from .regression.count import poisson, nbreg, ppmlhdfe
 from .neural_causal import tarnet, cfrnet, dragonnet, TARNet, CFRNet, DragonNet
 from .causal_discovery import notears, NOTEARS, pc_algorithm, PCAlgorithm, lingam, LiNGAMResult, ges, GESResult, fci, FCIResult
-from .tmle import tmle, TMLE, super_learner, SuperLearner
-from .policy_learning import policy_tree, PolicyTree, policy_value
+from .tmle import tmle, TMLE, super_learner, SuperLearner, ltmle, LTMLEResult
+from .policy_learning import policy_tree, PolicyTree, policy_value, direct_method, ips, snips, doubly_robust, OPEResult
 from .conformal_causal import conformal_cate, ConformalCATE
 from .bcf import bcf, BayesianCausalForest
 from .bunching import bunching, BunchingEstimator, notch, NotchResult
 from .matrix_completion import mc_panel, MCPanel
-from .dose_response import dose_response, DoseResponse
-from .bounds import lee_bounds, manski_bounds, BoundsResult, horowitz_manski, iv_bounds, oster_delta, selection_bounds, breakdown_frontier
-from .interference import spillover, SpilloverEstimator
-from .dtr import g_estimation, GEstimation
+from .dose_response import dose_response, DoseResponse, vcnet, scigan, VCNetResult
+from .bounds import lee_bounds, manski_bounds, BoundsResult, horowitz_manski, iv_bounds, oster_delta, selection_bounds, breakdown_frontier, balke_pearl, BalkePearlResult
+from .interference import spillover, SpilloverEstimator, network_exposure, NetworkExposureResult
+from .dtr import g_estimation, GEstimation, q_learning, QLearningResult, a_learning, ALearningResult, snmm, SNMMResult
 from .multi_treatment import multi_treatment, MultiTreatment
 from .robustness import spec_curve, SpecCurveResult, robustness_report, RobustnessResult, subgroup_analysis, SubgroupResult
 from .survey import svydesign, SurveyDesign, svymean, svytotal, svyglm, rake, linear_calibration
-from .dag import dag, DAG, dag_example, dag_examples, dag_example_positions, dag_simulate
+from .dag import (
+    dag, DAG, dag_example, dag_examples, dag_example_positions, dag_simulate,
+    identify, IdentificationResult,
+    rule1 as do_rule1, rule2 as do_rule2, rule3 as do_rule3,
+    apply_rules as do_calculus_apply, RuleCheck,
+    swig, SWIGGraph, SCM,
+)
+
+# === Target Trial Emulation (JAMA 2022 framework) ===
+from . import target_trial
+from .target_trial import (
+    protocol as target_trial_protocol,
+    emulate as target_trial_emulate,
+    clone_censor_weight,
+    immortal_time_check,
+    TargetTrialProtocol,
+    TargetTrialResult,
+    CloneCensorWeightResult,
+)
+# === Inverse probability of censoring weights ===
+from . import censoring
+from .censoring import ipcw, IPCWResult
 
 # === Canonical datasets (consolidated facade) ===
 from . import datasets
@@ -241,7 +267,7 @@
 # via fixest.wrapper._check_pyfixest, so top-level import never fails.
 from .fixest import feols, fepois, feglm, etable
 # Survival / Duration
-from .survival import cox, kaplan_meier, survreg, CoxResult, KMResult, logrank_test, cox_frailty, aft
+from .survival import cox, kaplan_meier, survreg, CoxResult, KMResult, logrank_test, cox_frailty, aft, causal_survival_forest, causal_survival, CausalSurvivalForestResult
 # Nonparametric
 from .nonparametric import lpoly, LPolyResult, kdensity, KDensityResult
 # Time Series (for causal inference)
@@ -252,6 +278,7 @@
     garch, GARCHResult,
     arima, ARIMAResult,
     bvar, BVARResult,
+    its, ITSResult,
 )
 # Experimental Design
 from .experimental import randomize, RandomizationResult, balance_check, BalanceResult, attrition_test, attrition_bounds, AttritionResult, optimal_design, OptimalDesignResult