🤖 Repo Assist here — I'm an automated AI assistant for this repository.

Activity for June 2026

Suggested Actions for Maintainer

Comprehensive list of all pending actions requiring maintainer attention.

• Review PR build(deps): bump github/gh-aw-actions/setup from 2fe53acc038ba01c3bbdc767d4b25df31ca5bdfc to ea222e359276c0702a5f5203547ff9d88d0ddd76 #1622 (pending): RA fix — initialise _variables_of_interest = [] before try block in CausalRefuter.__init__; prevents secondary AttributeError crash when estimate.estimator is not set — Review
• Review PR [Repo Assist] refactor(causal_graph): eliminate duplicated DOT/GML parsing in CausalGraph.__init__ #1621: RA refactor — eliminate ~35 lines of duplicated DOT/GML parsing in CausalGraph.__init__; delegate to build_graph_from_str() — Review
• Review PR [Repo Assist] deprecate(graph_learners): add DeprecationWarning to CausalModel.learn_graph() #1620: RA deprecation — add DeprecationWarning to CausalModel.learn_graph(), pointing users to causal-learn/dodiscover; partially closes Remove graph-learner API from DoWhy and point to causal-learn / dodiscover instead #1039 — Review
• Review PR Feature Request: Add Approximate Conley LTZ Sensitivity Analysis for IV Exclusion Violations #1613: contributor (abhiprd2000) — add IvExclusionSensitivityRefuter (Conley LTZ) for IV exclusion sensitivity analysis (closes Feature Request: Approximate Conley LTZ Sensitivity Analysis for IV Exclusion Violations #1612) — Review
• Review PR Fix conditional effect estimation crash on pandas >= 3.0 (include_groups removed) #1609: contributor (immu4989) — fix conditional effect estimation crash on pandas ≥ 3.0 (include_groups removed) — Review
• Review PR Fix categorical_treatment_model crash on scikit-learn >= 1.7 (multi_class removed) #1611: contributor (immu4989) — fix categorical_treatment_model crash on scikit-learn ≥ 1.7 (multi_class removed); updated to use OneVsRestClassifier — Review
• Review PR Clarify significance output and add dedicated significance_level parameter #1607: contributor (immu4989) — clarify significance output, add significance_level param (closes Bootstrap-based results are ambiguous #879) — Review
• Review PR [Repo Assist] ci: add fail-fast: false to matrix strategies #1619: RA CI — add fail-fast: false to CI matrix strategies (ci.yml, nightly-tests.yml, ci-install.yml) — requires human push (workflow files) — View
• Review PR [Repo Assist] docs(identification): expand frontdoor and IV criterion docs from stubs #1618: RA docs — expand frontdoor.rst and instrumental_variable.rst from stubs (closes Add short descriptions including code examples for each identification method #531 partially) — Review
• Review PR [Repo Assist] fix(api): accept num_quantiles_to_discretize_cont_cols as direct kwarg in estimate_effect() #1615: RA fix — accept num_quantiles_to_discretize_cont_cols as direct kwarg in estimate_effect() (closes effect_modifiers #767, estimate_effect() got an unexpected keyword argument 'num_quantiles_to_discretize_cont_cols' #1035) — Review
• Review PR [Repo Assist] fix(econml): raise helpful ValueError when X=None for estimators requiring effect_modifiers #1614: RA fix — raise helpful ValueError when X=None for EconML estimators requiring effect_modifiers (closes EconML CausalForestDML Estimation error #801) — Review
• Review PR [Repo Assist] fix(distance_matching): route Minkowski p param as first-class NearestNeighbors arg #1603: RA fix — route Minkowski p as first-class NearestNeighbors arg; supersedes [Repo Assist] fix: route distance metric params through metric_params= in NearestNeighbors (fixes #1390) #1577; notebook CI failure is infrastructure-related (all unit tests pass) — Review
• Review PR [Repo Assist] perf(falsify): replace N×nx.ancestors calls with transitive_closure_dag in validate_pd #1606: RA perf — replace N×nx.ancestors calls with transitive_closure_dag in validate_pd (falsify.py) — Review
• Review PR Fix ConfounderDistributionInterpreter crash on categorical confounders #1586: contributor (immu4989) — fix ConfounderDistributionInterpreter crash on categorical confounders (closes #1588) — Review
• Review PR [Repo Assist] improve(gcm): add random_seed parameter to interventional_samples, counterfactual_samples, and average_causal_effect #1549: RA improve — add random_seed to interventional_samples, counterfactual_samples, average_causal_effect — ✅ all CI passing — Review
• Review PR [Repo Assist] perf(gcm): replace O(N×K) nx.ancestors loop with O(K) nx.descendants in _get_nodes_affected_by_intervention #1600: RA perf — replace O(N×K) nx.ancestors loop with O(K) nx.descendants; ✅ all CI passing — Review
• Review PR [Repo Assist] improve: warn users when treatment/outcome columns contain NaN values #1599: RA improve — warn UserWarning when treatment/outcome columns contain NaN (updated per @emrekiciman feedback) — Review
• Review PR [Repo Assist] docs: add academic citations to causal estimator docstrings #1598: RA docs — add academic citations to 7 causal estimator docstrings (closes Add citations to papers for identification and estimation methods #575) — Review
• Review PR [Repo Assist] perf(gcm): skip encoder fitting in select_model when feature count exceeds threshold #1596: RA perf — skip encoder fitting in select_model() when X.shape[1] > 5 — Review
• Review PR [Repo Assist] test(utils): add Encoders class tests and one_hot_encode edge cases #1580: RA test — adds 11 Encoders class tests + 3 one_hot_encode edge-case tests — Review
• Review PR [Repo Assist] perf(gcm): avoid pre-allocated DataFrame in compute_data_from_noise and compute_noise_from_data #1561: RA perf — avoid pre-allocated DataFrame in _noise.py — Review
• Review PR [Repo Assist] fix: preserve mediation confounders and no_directed_path in IdentifiedEstimand.__deepcopy__ #1548: RA fix — preserve mediation_first_stage_confounders in IdentifiedEstimand.__deepcopy__ — Review
• Review PR [Repo Assist] fix(cit): correct conditional_MI entropy computation for discrete variables #1547: RA fix — correct conditional_MI entropy computation; rebased 2026-06-23 — Review
• Review PR [Repo Assist] perf(gcm): avoid redundant get_ordered_predecessors calls and pre-allocated DataFrame in fitting_sampling #1544: RA perf — avoid redundant get_ordered_predecessors calls — Review
• Review PR [Repo Assist] improve: replace generic Exception with specific exception types in remaining modules #1543: RA improve — replace generic Exception with specific types — Review
• Review PR [Repo Assist] feat(gcm): add optional nodes filter to anomaly_scores() #1540: RA feat — add optional nodes filter to gcm.anomaly_scores() — Review
• Review PR [Repo Assist] docs: improve refutation documentation with p-value interpretation guide #1535: RA docs — improve refutation docs with p-value interpretation table — Review
• Review PR [Repo Assist] feat(datasets): add return_all_coefficients parameter to linear_dataset #1530: RA feat — add return_all_coefficients to linear_dataset() (closes Causal Effect of Common Causes to Outcome #466) — Review
• Review PR [Repo Assist] fix: make refuter significance tests robust to array-valued estimates #1529: RA fix — perform_bootstrap_test robust to array-valued estimates — Review
• Review PR [Repo Assist] feat(gcm): expose plotting functions at dowhy.gcm level #1528: RA feat — expose plotting functions at dowhy.gcm top level (closes Move gcm plotting utility one level up #586) — Review
• Review PR [Repo Assist] fix(cit): correct conditional_MI column indexing to use Series instead of DataFrame #1522: RA fix — correct conditional_MI column indexing (consolidate with or close in favour of [Repo Assist] fix(cit): correct conditional_MI entropy computation for discrete variables #1547) — Review
• Review PR [Repo Assist] docs: fix TwoStageRegressionEstimator docstring and fix construct_symbolic_estimator error message #1516: RA docs — fix TwoStageRegressionEstimator docstring — Review
• Review PR [Repo Assist] fix: accept num_quantiles_to_discretize_cont_cols and sample_size_fraction directly in estimate_effect() #1515: RA fix — accept extra params in estimate_effect() — Review
• Review PR [Repo Assist] improve: use .get() with defaults in refute_graph independence_test dict #1514: RA improve — use .get() in refute_graph independence_test dict — Review
• Review PR [Repo Assist] fix: respect treatment_value/control_value in PropensityScoreStratificationEstimator #1512: RA fix — respect treatment_value/control_value in PropensityScoreStratificationEstimator — Review
• Review PR [Repo Assist] fix: implement asymptotic CI/SE via Delta method for LinearRegressionEstimator with effect modifiers #1432: RA fix — asymptotic CI/SE via Delta method for LinearRegressionEstimator (closes Add asymptotic confidence intervals for average treatment effect for linear regression with effect modifiers #336) — Review
• Review PR feat: Add z-identifiability via surrogate experiments as identification method in identify_ate_effect #1531: contributor — add z-identifiability via surrogate experiments — Review
• Review PR chore: improve dowhy maintenance path #1566: contributor — improve dowhy maintenance path — Review
• Review PR build(deps): bump actions/checkout from 5 to 7 #1601: Dependabot — bump actions/checkout from 5 to 7 — Review
• Review PR [Repo Assist] fix: preserve DataFrame index in one_hot_encode to fix data_subset_refuter bug with categorical columns #1408: RA fix — preserve DataFrame index in one_hot_encode to fix data_subset_refuter bug with categorical columns — Review
• Review PR [Repo Assist] test: add tests for RandomCommonCause refuter #1413: RA test — add tests for RandomCommonCause refuter — Review
• Check comment on PR Feature Request: Add Approximate Conley LTZ Sensitivity Analysis for IV Exclusion Violations #1613: RA reviewed PR — noted API discoverability gap (method_name not documented), constructor pattern, missing docstring, has_intercept fragility — View
• Check comment on issue graph learning and R depedency on CDT #1021: RA noted causal discovery notebook already uses causal-learn; issue may be closeable after PR [Repo Assist] deprecate(graph_learners): add DeprecationWarning to CausalModel.learn_graph() #1620 merges — View
• Check comment on issue Move test_significance from estimator property to estimate_effect parameter? #871: RA analyzed current test_significance code state and proposed backward-compatible path — View
• Check comment on issue Add links to causal inference learning resources #534: RA added list of causal inference learning resources to add to docs — View
• Check comment on issue Github action that runs notebooks #830: RA noted notebooks already run in CI; issue may be closeable — View
• Check comment on PR [Repo Assist] fix(distance_matching): route Minkowski p param as first-class NearestNeighbors arg #1603: RA noted the notebook CI failure is infrastructure-related (all unit tests pass) — View
• Check comment Check input data for NAN values before the estimation is done #827: RA updated PR [Repo Assist] improve: warn users when treatment/outcome columns contain NaN values #1599 warning text per maintainer feedback — View
• Check comment Linear dataset functionality and parameters #1045: RA noted PR [Repo Assist] feat(datasets): add return_all_coefficients parameter to linear_dataset #1530 adds return_all_coefficients; workarounds for range constraints — View
• Check comment Errors of Notebook: The Causal Story Behind Hotel Booking Cancellations #198: RA suggested closing (issue resolved since 2021) — View
• Close PR [Repo Assist] fix: route distance metric params through metric_params= in NearestNeighbors (fixes #1390) #1577: superseded by [Repo Assist] fix(distance_matching): route Minkowski p param as first-class NearestNeighbors arg #1603 (same fix, cleaner) — Close
• Close issue PC: background_knowledge not strictly enforced (tiers & forbidden_edges); results vary by uc_rule / uc_priority #1343: PC background_knowledge is a causal-learn issue, not DoWhy — View
• Close issue Errors of Notebook: The Causal Story Behind Hotel Booking Cancellations #198: fully resolved since 2021 — View
• Close issue [gh-aw] Protected Files: [Repo Assist] fix(cit): correct conditional_MI entropy computation for discrete variables #1602: [gh-aw] Protected Files notification — safe to close — View
• Close issue [gh-aw] Protected Files: [Repo Assist] fix(cit): correct conditional_MI entropy computation for discrete variables #1604: [gh-aw] Protected Files notification — safe to close — View
• Close issue [gh-aw] Protected Files: [Repo Assist] perf(gcm): replace O(N×K) nx.ancestors loop with O(K) nx.descendants in _get_nodes_affected_by_intervention #1605: [gh-aw] Protected Files notification — safe to close — View
• Close issue [gh-aw] Protected Files: [Repo Assist] improve: warn users when treatment/outcome columns contain NaN values #1616: [gh-aw] Protected Files notification — safe to close — View
• Close issue [gh-aw] Protected Files: [Repo Assist] perf(gcm): replace O(N×K) nx.ancestors loop with O(K) nx.descendants in _get_nodes_affected_by_intervention #1617: [gh-aw] Protected Files notification — safe to close — View

Future Work for Repo Assist

• Many RA PRs pending CI approval — maintainer needs to approve CI runs for bot PRs.
• Continue rebasing stale PRs as main advances.
• PR [Repo Assist] improve: warn users when treatment/outcome columns contain NaN values #1599 NaN warning updated per @emrekiciman feedback — ready for review.
• Issue Move test_significance from estimator property to estimate_effect parameter? #871 (test_significance architecture): RA commented with implementation plan — consider implementing if maintainers agree.
• Issue graph learning and R depedency on CDT #1021 causal discovery notebook already updated to use causal-learn — consider closing once PR [Repo Assist] deprecate(graph_learners): add DeprecationWarning to CausalModel.learn_graph() #1620 merges.

Run History

2026-06-29 14:05 UTC — Run

• 💬 Task 2: Reviewed PR Feature Request: Add Approximate Conley LTZ Sensitivity Analysis for IV Exclusion Violations #1613 — technical feedback on Conley LTZ refuter (API discoverability, constructor pattern, missing docstring, has_intercept fragility)
• 🔧 Task 3: Created PR build(deps): bump github/gh-aw-actions/setup from 2fe53acc038ba01c3bbdc767d4b25df31ca5bdfc to ea222e359276c0702a5f5203547ff9d88d0ddd76 #1622 (pending) — fix _variables_of_interest initialisation in CausalRefuter.__init__; prevents secondary AttributeError when estimate.estimator is not set

2026-06-29 01:33 UTC — Run

• 💬 Task 2: Commented on graph learning and R depedency on CDT #1021 — noted causal discovery notebook already uses causal-learn; issue may be closeable after PR [Repo Assist] deprecate(graph_learners): add DeprecationWarning to CausalModel.learn_graph() #1620 merges
• 🔧 Task 5: Created PR [Repo Assist] refactor(causal_graph): eliminate duplicated DOT/GML parsing in CausalGraph.__init__ #1621 — refactor: eliminate ~35 lines of duplicated DOT/GML parsing in CausalGraph.__init__; delegate to build_graph_from_str()

2026-06-28 13:07 UTC — Run

• 💬 Task 2: Commented on Move test_significance from estimator property to estimate_effect parameter? #871 — analyzed current test_significance code state; proposed backward-compatible path forward allowing per-call override
• 🔧 Task 4: Created PR [Repo Assist] deprecate(graph_learners): add DeprecationWarning to CausalModel.learn_graph() #1620 — add DeprecationWarning to CausalModel.learn_graph() pointing users to causal-learn/dodiscover; partially closes #1039

2026-06-28 01:32 UTC — Run

• 🔧 Task 4: Created PR [Repo Assist] ci: add fail-fast: false to matrix strategies #1619 — add fail-fast: false to CI matrix strategies (ci.yml, nightly-tests.yml, ci-install.yml)
• 💬 Task 2: Commented on Errors of Notebook: The Causal Story Behind Hotel Booking Cancellations #198 — noted issue fully resolved since 2021, suggested closing

2026-06-27 01:26 UTC — Run

• 💬 Task 2: Commented on Add links to causal inference learning resources #534 — added list of recommended causal inference learning resources for the docs stub
• 💬 Task 2: Commented on Github action that runs notebooks #830 — noted notebooks already run in CI (pytest + nbconvert), suggested issue could be closed
• 💬 Task 6: Commented on PR [Repo Assist] fix(distance_matching): route Minkowski p param as first-class NearestNeighbors arg #1603 — documented that the single CI failure (notebook test) is infrastructure-related; all unit tests pass

2026-06-26 13:25 UTC — Run

• 💬 Task 2: Commented on Linear dataset functionality and parameters #1045 — noted PR [Repo Assist] feat(datasets): add return_all_coefficients parameter to linear_dataset #1530 (return_all_coefficients), range-constraint workarounds, and GCM custom mechanism approach
• 🔧 Task 5: Created docs PR [Repo Assist] docs(identification): expand frontdoor and IV criterion docs from stubs #1618 — expand frontdoor.rst and instrumental_variable.rst from stubs with math, intuition, code examples (closes #531 partially)

2026-06-26 01:31 UTC — Run

• 🔄 Task 6: Rebased PR [Repo Assist] perf(gcm): replace O(N×K) nx.ancestors loop with O(K) nx.descendants in _get_nodes_affected_by_intervention #1600 onto main (da3cd56) — nx.descendants perf fix now current; 15 test_whatif.py tests pass
• 🔍 Task 2: Scanned all open issues — all already have Repo Assist comments from previous runs

2026-06-25 13:28 UTC — Run

• 🔄 Task 3: Updated PR [Repo Assist] improve: warn users when treatment/outcome columns contain NaN values #1599 — revised NaN warning text (removed dropna() suggestion per @emrekiciman feedback; add bias-aware message); fixed unused import in test (flake8 F401)
• 🔍 Task 2: Scanned all open issues — all already have Repo Assist comments from previous runs

2026-06-25 01:29 UTC — Run

• 🔧 Task 3: Created PR [Repo Assist] fix(api): accept num_quantiles_to_discretize_cont_cols as direct kwarg in estimate_effect() #1615 — accept num_quantiles_to_discretize_cont_cols as direct kwarg in estimate_effect() (closes #767, estimate_effect() got an unexpected keyword argument 'num_quantiles_to_discretize_cont_cols' #1035); 4 tests pass
• 💬 Task 2: Commented on PC: background_knowledge not strictly enforced (tiers & forbidden_edges); results vary by uc_rule / uc_priority #1343 — PC background_knowledge API is a causal-learn issue, not DoWhy

2026-06-24 13:29 UTC — Run

• 🔧 Task 3: Created PR [Repo Assist] fix(econml): raise helpful ValueError when X=None for estimators requiring effect_modifiers #1614 for EconML CausalForestDML Estimation error #801 — improve EconML X=None error message; adds actionable hint to specify effect_modifiers; 65 tests pass
• 💬 Task 2: Commented on Feature Request: Approximate Conley LTZ Sensitivity Analysis for IV Exclusion Violations #1612 — design review notes for PR Feature Request: Add Approximate Conley LTZ Sensitivity Analysis for IV Exclusion Violations #1613 (IV exclusion sensitivity refuter)
• 💬 Task 2: Commented on PR Clarify significance output and add dedicated significance_level parameter #1607 — flagged fragile test assertion ("significant" substring) and {:.2g} rounding concern from Copilot review

2026-06-24 01:23 UTC — Run

• 🏷️ Labelled Conditional effect estimation crashes on pandas >= 3.0 (include_groups=True is no longer allowed) #1608 with bug — pandas ≥ 3.0 include_groups crash
• 🏷️ Labelled categorical_treatment_model crashes on scikit-learn >= 1.7 (LogisticRegression multi_class removed) #1610 with bug — scikit-learn ≥ 1.7 multi_class crash
• 💬 Task 2: Commented on Conditional effect estimation crashes on pandas >= 3.0 (include_groups=True is no longer allowed) #1608 — acknowledged root cause, endorsed PR Fix conditional effect estimation crash on pandas >= 3.0 (include_groups removed) #1609 fix
• 💬 Task 2: Commented on categorical_treatment_model crashes on scikit-learn >= 1.7 (LogisticRegression multi_class removed) #1610 — acknowledged root cause, endorsed PR Fix categorical_treatment_model crash on scikit-learn >= 1.7 (multi_class removed) #1611 fix
• 💬 Task 6: Commented on PR [Repo Assist] fix: route distance metric params through metric_params= in NearestNeighbors (fixes #1390) #1577 — flagged as superseded by [Repo Assist] fix(distance_matching): route Minkowski p param as first-class NearestNeighbors arg #1603, recommended close

2026-06-23 13:37 UTC — Run

• 🔧 Task 8: Created PR [Repo Assist] perf(falsify): replace N×nx.ancestors calls with transitive_closure_dag in validate_pd #1606 — replace N×nx.ancestors calls with transitive_closure_dag in validate_pd; adjacent_only=True short-circuits to causal_graph.edges()
• 💬 Task 2: Commented on Check input data for NAN values before the estimation is done #827 — acknowledged maintainer's NaN-bias caution; flagged that PR [Repo Assist] improve: warn users when treatment/outcome columns contain NaN values #1599 warning text should point to proper missing-data handling, not dropna()

2026-06-23 01:26 UTC — Run

• 🔄 Task 6: Rebased PR [Repo Assist] fix(cit): correct conditional_MI entropy computation for discrete variables #1547 onto main (d955cee) — conditional_MI entropy fix now current; fresh CI pending
• 🔄 Task 6: Rebased PR [Repo Assist] perf(gcm): replace O(N×K) nx.ancestors loop with O(K) nx.descendants in _get_nodes_affected_by_intervention #1600 onto main (d955cee) — nx.descendants perf fix now current

2026-06-22 14:37 UTC — Run

• 🔧 Task 3: Created PR [Repo Assist] fix(distance_matching): route Minkowski p param as first-class NearestNeighbors arg #1603 — fix DistanceMatchingEstimator Minkowski p param (supersedes [Repo Assist] fix: route distance metric params through metric_params= in NearestNeighbors (fixes #1390) #1577, closes #1390)
• 🔄 Task 6: Rebased PR [Repo Assist] fix(cit): correct conditional_MI entropy computation for discrete variables #1547 onto main

2026-06-22 01:41 UTC — Run

• 💬 Task 2: Commented on PC: background_knowledge not strictly enforced (tiers & forbidden_edges); results vary by uc_rule / uc_priority #1343 — PC/background_knowledge is a causal-learn issue
• 🔧 Task 8: Created PR [Repo Assist] perf(gcm): replace O(N×K) nx.ancestors loop with O(K) nx.descendants in _get_nodes_affected_by_intervention #1600 — replace O(N×K) nx.ancestors with O(K) nx.descendants; 15 tests pass

2026-06-21 13:22 UTC — Run

• 🔧 Task 5: Created PR [Repo Assist] improve: warn users when treatment/outcome columns contain NaN values #1599 — warn on NaN in treatment/outcome before fitting
• 🔧 Task 10: Created PR [Repo Assist] docs: add academic citations to causal estimator docstrings #1598 — add academic citations to 7 estimator docstrings (closes #575)
• 💬 Task 10: Commented on Case-study notebooks for practical usage of DoWhy #533, Add citations to papers for identification and estimation methods #575

Earlier runs (2026-06-01 through 2026-06-21)

• Created PRs: [Repo Assist] fix: preserve DataFrame index in one_hot_encode to fix data_subset_refuter bug with categorical columns #1408, [Repo Assist] test: add tests for RandomCommonCause refuter #1413, [Repo Assist] fix: implement asymptotic CI/SE via Delta method for LinearRegressionEstimator with effect modifiers #1432, [Repo Assist] test: add unit-test coverage for GraphRefuter and GraphRefutation #1479, [Repo Assist] fix: warn when treatment/outcome columns contain NaN values (closes #827) #1481, [Repo Assist] fix: identify all parallel mediators in identify_mediation() #1483–Bump notebook from 7.4.4 to 7.5.6 #1492, [Repo Assist] fix: do not qcut boolean effect modifiers in estimate_conditional_effects #1499–[Repo Assist] docs: fix TwoStageRegressionEstimator docstring and fix construct_symbolic_estimator error message #1516, [Repo Assist] fix(cit): correct conditional_MI column indexing to use Series instead of DataFrame #1522, [Repo Assist] feat(gcm): expose plotting functions at dowhy.gcm level #1528–[Repo Assist] feat(datasets): add return_all_coefficients parameter to linear_dataset #1530, [Repo Assist] docs: improve refutation documentation with p-value interpretation guide #1535, [Repo Assist] feat(gcm): add optional nodes filter to anomaly_scores() #1540, [Repo Assist] improve: replace generic Exception with specific exception types in remaining modules #1543–[Repo Assist] perf(gcm): avoid redundant get_ordered_predecessors calls and pre-allocated DataFrame in fitting_sampling #1544, [Repo Assist] fix(cit): correct conditional_MI entropy computation for discrete variables #1547–[Repo Assist] improve(gcm): add random_seed parameter to interventional_samples, counterfactual_samples, and average_causal_effect #1549, [Repo Assist] perf(gcm): avoid pre-allocated DataFrame in compute_data_from_noise and compute_noise_from_data #1561, [Repo Assist] test(utils): add Encoders class tests and one_hot_encode edge cases #1580
• Commented on 40+ issues including: DoSampler: Fails when you Add a Graph to the Causal Model #83, Mediation Analysis key error #214, arrays used as indices must be of integer (or boolean) type #225, Mediation analysis: how to choose mediators? #249, effect_estimate different using do operator and _estimate_effect function #309, Include conditional effects, p-value and confidence intervals for DML #325, Add asymptotic confidence intervals for average treatment effect for linear regression with effect modifiers #336, AttributeError: 'CausalEstimate' object has no attribute '_estimator_object' #357, Observed Confounder Strength for Binary Y or T #381, Error in Conditional Effect Estimation in backdoor.linear_regression_estimator #401, regression adjusted estimator for propensity_score_weighting #407, Adding model validation after user specifies model #412–'fit_estimator' argument in 'estimate_effect' function of 'CausalModel' object #414, Algorithms for efficient adjustment sets #464, Causal Effect of Common Causes to Outcome #466, Categorical Covariates in Forest models #471, Evaluating fitted built-in estimators on new dataset #484, Add video/screenshot guide for creating graph in DAGitty and importing into DoWhy #529–Add short descriptions with a code example for each of the estimation methods #530, Case-study notebooks for practical usage of DoWhy #533, Pass Solver and n_iter parameters in the model estimation step #560, Add citations to papers for identification and estimation methods #575, Move gcm plotting utility one level up #586–Move independence test package one level up #587, Support selection bias for identification (selection-backdoor criterion) #603, Prototyping panel data support for causal effect inference #635, Update dowhy_simple_example to clarify that dataframe columns are used by identify_effect() to indicate features are observed  #638, Add support for non-singleton treatment/outcome sets to "has_directed_path" method #654, Enhancement request to warn if columns in data frame are not included in the causal graph #670, Add support for logistic regression in mediation functionality #688, Include percentage increase as an option for effect output #740, effect_modifiers #767, EconML CausalForestDML Estimation error #801, Check input data for NAN values before the estimation is done #827, cannot import name 'CausalIdentifier' from 'dowhy.causal_identifier' #831, Increase the default number of simulations #841, Vectorize matching and ATE computation to speed up computation #935, Error handling column names in CIT as used by CausalModel.graph_refute method #949, estimate_effect() got an unexpected keyword argument 'num_quantiles_to_discretize_cont_cols' #1035–ValueError: The truth value of a Series is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all(). #1036, Linear dataset functionality and parameters #1045, numpy has no attribute 'long' #1052, Causal Graph not provided. DoWhy will construct a graph based on data inputs. #1125, Support polars data frames #1151, numpy.dual is dropped but it still occurs in dowhy #1181, No Backdoor Path Available #1188, model.estimate_effect and model.refute_astimate throws 'A column-vector y was passed ...' error #1212, Error with Do operator.  #1240–Counterfactual Samples giving invalid values of effect in dowhy gcm #1241, In gcm, allow building a model in presence of missing values if an appropriate estimator is selected #1300, Interventional samples gives inconsistent results. #1307, Warnings using causal.model #1316, PC: background_knowledge not strictly enforced (tiers & forbidden_edges); results vary by uc_rule / uc_priority #1343, Error in the tutorial of "Finding optimal adjustment sets" #1360, Add an option to specify only target and causal dependency outlier metrics during inference for Do-Why #1364–The causal structure returned by causal-learn is used in DoWhy. #1365, Yet another stundet coursework #1375, Issue with distance_metric_params for dowhy.causal_estimators.distance_matching_estimator #1390
• Labelled issues: Conditional effect estimation crashes on pandas >= 3.0 (include_groups=True is no longer allowed) #1608 (bug), categorical_treatment_model crashes on scikit-learn >= 1.7 (LogisticRegression multi_class removed) #1610 (bug)

Generated by 🌈 Repo Assist, see workflow run. Learn more.

Generated by 🌈 Repo Assist, see workflow run. Learn more.

To install this agentic workflow, run

gh aw add githubnext/agentics/workflows/repo-assist.md@11c9a2c442e519ff2b427bf58679f5a525353f76

---

2026-06-30 01:35 UTC — Run

• 🔄 Task 6: Rebased PR [Repo Assist] fix(cit): correct conditional_MI entropy computation for discrete variables #1547 (conditional_MI fix) onto main (da3cd56) — also improved docstring for conditional_MI; CI run pending
• 🔄 Task 6: Rebased PR [Repo Assist] feat(gcm): add optional nodes filter to anomaly_scores() #1540 (anomaly_scores nodes filter) onto main — added node validation ValueError + regression test; CI run pending
• 💬 Task 2: Commented on Add examples from "Causal Inference and Discovery in Python" to the tests to ensure backwards compatibility #970 — pointed out test structure conventions (@pytest.mark.advanced, tests/causal_inference_discovery_book/), encouraged @rahulbshrestha and @AlxndrMlk to restart contribution

Generated by 🌈 Repo Assist, see workflow run. Learn more.

To install this agentic workflow, run

gh aw add githubnext/agentics/workflows/repo-assist.md@11c9a2c442e519ff2b427bf58679f5a525353f76