From 83790071f7fe1492a154a024ac574ca988bfc4b5 Mon Sep 17 00:00:00 2001
From: "github-actions[bot]"
 <41898282+github-actions[bot]@users.noreply.github.com>
Date: Sat, 30 May 2026 13:22:18 +0000
Subject: [PATCH] fix: raise ValueError in estimate_effect() when identified
 estimand is None

Previously, calling estimate_effect() with a method whose estimand is None
(e.g. 'iv.instrumental_variable' when no instruments are in the graph) would
silently return a CausalEstimate with value=None, with only a logger.error
message. Callers had no programmatic indication that identification failed.

This replaces the silent return with a ValueError that includes a descriptive
message explaining which identifier failed and why, matching the contract
expected by users (per issue #1551).

Closes #1551

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Signed-off-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
---
 dowhy/causal_estimator.py                     |  7 ++---
 .../test_instrumental_variable_estimator.py   | 27 +++++++++++++++++++
 2 files changed, 31 insertions(+), 3 deletions(-)

diff --git a/dowhy/causal_estimator.py b/dowhy/causal_estimator.py
index 23de0587ac..895a13cd7f 100755
--- a/dowhy/causal_estimator.py
+++ b/dowhy/causal_estimator.py
@@ -744,9 +744,10 @@ def estimate_effect(
         )
     # Check if estimator's target estimand is identified
     elif identified_estimand.estimands[identifier_name] is None:
-        logger.error("No valid identified estimand available.")
-        return CausalEstimate(
-            None, None, None, None, None, None, control_value=control_value, treatment_value=treatment_value
+        raise ValueError(
+            f"No valid identified estimand for '{identifier_name}'. "
+            "Ensure that the identification step succeeded for this estimator method "
+            "(e.g. the graph must contain valid instruments for 'iv.instrumental_variable')."
         )
 
     if fit_estimator:
diff --git a/tests/causal_estimators/test_instrumental_variable_estimator.py b/tests/causal_estimators/test_instrumental_variable_estimator.py
index b6705d3ccc..2e9a8df540 100755
--- a/tests/causal_estimators/test_instrumental_variable_estimator.py
+++ b/tests/causal_estimators/test_instrumental_variable_estimator.py
@@ -3,6 +3,8 @@
 import pytest
 from pytest import mark
 
+import dowhy.datasets
+from dowhy import CausalModel
 from dowhy.causal_estimators.instrumental_variable_estimator import InstrumentalVariableEstimator
 
 from .base import SimpleEstimator
@@ -78,3 +80,28 @@ def test_average_treatment_effect(
         cfg["num_instruments"] = 0
         with pytest.raises(ValueError):
             estimator_tester.average_treatment_effect_test(**cfg)
+
+    def test_estimate_effect_raises_when_iv_estimand_is_none(self):
+        """Regression test for #1551: estimate_effect() must raise ValueError
+        instead of silently returning CausalEstimate(value=None) when the
+        identified estimand for 'iv' is None (i.e. no instruments in the graph).
+        """
+        data = dowhy.datasets.linear_dataset(
+            beta=10,
+            num_common_causes=2,
+            num_instruments=0,
+            num_samples=500,
+            treatment_is_binary=True,
+        )
+        model = CausalModel(
+            data=data["df"],
+            treatment=data["treatment_name"],
+            outcome=data["outcome_name"],
+            graph=data["dot_graph"],
+        )
+        identified_estimand = model.identify_effect(proceed_when_unidentifiable=True)
+        assert identified_estimand.estimands.get("iv") is None, (
+            "Precondition: iv estimand must be None when num_instruments=0"
+        )
+        with pytest.raises(ValueError):
+            model.estimate_effect(identified_estimand, method_name="iv.instrumental_variable")