From 718dcde32829efad941f97171201f12ccc86bea1 Mon Sep 17 00:00:00 2001 From: Amit Sharma Date: Sat, 27 Jul 2019 19:12:41 +0530 Subject: [PATCH] fixed bug in IV estimator for continuous IV --- docs/source/dowhy_estimation_methods.ipynb | 229 ++++++++---------- .../instrumental_variable_estimator.py | 2 +- 2 files changed, 100 insertions(+), 131 deletions(-) diff --git a/docs/source/dowhy_estimation_methods.ipynb b/docs/source/dowhy_estimation_methods.ipynb index 027e88a45..1f0c3f0d5 100644 --- a/docs/source/dowhy_estimation_methods.ipynb +++ b/docs/source/dowhy_estimation_methods.ipynb @@ -109,22 +109,7 @@ "cell_type": "code", "execution_count": 5, "metadata": {}, - "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "WARNING:dowhy.causal_graph:Warning: Pygraphviz cannot be loaded. Check that graphviz and pygraphviz are installed.\n", - "INFO:dowhy.causal_graph:Using Matplotlib for plotting\n", - "/home/amit/virtualenvs/python37/lib/python3.7/site-packages/networkx/drawing/nx_pylab.py:579: MatplotlibDeprecationWarning: \n", - "The iterable function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use np.iterable instead.\n", - " if not cb.iterable(width):\n", - "/home/amit/virtualenvs/python37/lib/python3.7/site-packages/networkx/drawing/nx_pylab.py:676: MatplotlibDeprecationWarning: \n", - "The iterable function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use np.iterable instead.\n", - " if cb.iterable(node_size): # many node sizes\n" - ] - } - ], + "outputs": [], "source": [ "model.view_model()" ] @@ -136,7 +121,7 @@ "outputs": [ { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "" ] @@ -166,7 +151,7 @@ "name": "stderr", "output_type": "stream", "text": [ - "INFO:dowhy.causal_identifier:Common causes of treatment and outcome:['Unobserved Confounders', 'X0', 'X2', 'X3', 'X1', 'X4']\n", + "INFO:dowhy.causal_identifier:Common causes of treatment and outcome:['X0', 'X1', 'X4', 'X3', 'X2', 'Unobserved Confounders']\n", "WARNING:dowhy.causal_identifier:There are unobserved common causes. Causal effect cannot be identified.\n" ] }, @@ -181,7 +166,7 @@ "name": "stderr", "output_type": "stream", "text": [ - "INFO:dowhy.causal_identifier:Instrumental variables for treatment and outcome:['Z1', 'Z0']\n" + "INFO:dowhy.causal_identifier:Instrumental variables for treatment and outcome:['Z0', 'Z1']\n" ] }, { @@ -190,18 +175,18 @@ "text": [ "Estimand type: ate\n", "### Estimand : 1\n", + "Estimand name: iv\n", + "Estimand expression:\n", + "Expectation(Derivative(y, Z0)/Derivative(v, Z0))\n", + "Estimand assumption 1, Exclusion: If we remove {Z0,Z1}→v, then ¬(Z0,Z1→y)\n", + "Estimand assumption 2, As-if-random: If U→→y then ¬(U →→Z0,Z1)\n", + "### Estimand : 2\n", "Estimand name: backdoor\n", "Estimand expression:\n", "d \n", - "──(Expectation(y|X0,X2,X3,X1,X4))\n", + "──(Expectation(y|X0,X1,X4,X3,X2))\n", "dv \n", - "Estimand assumption 1, Unconfoundedness: If U→v and U→y then P(y|v,X0,X2,X3,X1,X4,U) = P(y|v,X0,X2,X3,X1,X4)\n", - "### Estimand : 2\n", - "Estimand name: iv\n", - "Estimand expression:\n", - "Expectation(Derivative(y, Z1)/Derivative(v, Z1))\n", - "Estimand assumption 1, As-if-random: If U→→y then ¬(U →→Z1,Z0)\n", - "Estimand assumption 2, Exclusion: If we remove {Z1,Z0}→v, then ¬(Z1,Z0→y)\n", + "Estimand assumption 1, Unconfoundedness: If U→v and U→y then P(y|v,X0,X1,X4,X3,X2,U) = P(y|v,X0,X1,X4,X3,X2)\n", "\n" ] } @@ -230,7 +215,7 @@ "output_type": "stream", "text": [ "INFO:dowhy.causal_estimator:INFO: Using Linear Regression Estimator\n", - "INFO:dowhy.causal_estimator:b: y~v+X0+X2+X3+X1+X4\n" + "INFO:dowhy.causal_estimator:b: y~v+X0+X1+X4+X3+X2\n" ] }, { @@ -242,28 +227,28 @@ "## Target estimand\n", "Estimand type: ate\n", "### Estimand : 1\n", + "Estimand name: iv\n", + "Estimand expression:\n", + "Expectation(Derivative(y, Z0)/Derivative(v, Z0))\n", + "Estimand assumption 1, Exclusion: If we remove {Z0,Z1}→v, then ¬(Z0,Z1→y)\n", + "Estimand assumption 2, As-if-random: If U→→y then ¬(U →→Z0,Z1)\n", + "### Estimand : 2\n", "Estimand name: backdoor\n", "Estimand expression:\n", "d \n", - "──(Expectation(y|X0,X2,X3,X1,X4))\n", + "──(Expectation(y|X0,X1,X4,X3,X2))\n", "dv \n", - "Estimand assumption 1, Unconfoundedness: If U→v and U→y then P(y|v,X0,X2,X3,X1,X4,U) = P(y|v,X0,X2,X3,X1,X4)\n", - "### Estimand : 2\n", - "Estimand name: iv\n", - "Estimand expression:\n", - "Expectation(Derivative(y, Z1)/Derivative(v, Z1))\n", - "Estimand assumption 1, As-if-random: If U→→y then ¬(U →→Z1,Z0)\n", - "Estimand assumption 2, Exclusion: If we remove {Z1,Z0}→v, then ¬(Z1,Z0→y)\n", + "Estimand assumption 1, Unconfoundedness: If U→v and U→y then P(y|v,X0,X1,X4,X3,X2,U) = P(y|v,X0,X1,X4,X3,X2)\n", "\n", "## Realized estimand\n", - "b: y~v+X0+X2+X3+X1+X4\n", + "b: y~v+X0+X1+X4+X3+X2\n", "## Estimate\n", - "Value: 10.00000000000001\n", + "Value: 9.999999999999075\n", "\n", "## Statistical Significance\n", "p-value: <0.001\n", "\n", - "Causal Estimate is 10.00000000000001\n" + "Causal Estimate is 10.0\n" ] } ], @@ -294,7 +279,7 @@ "output_type": "stream", "text": [ "INFO:dowhy.causal_estimator:INFO: Using Propensity Score Stratification Estimator\n", - "INFO:dowhy.causal_estimator:b: y~v+X0+X2+X3+X1+X4\n" + "INFO:dowhy.causal_estimator:b: y~v+X0+X1+X4+X3+X2\n" ] }, { @@ -306,33 +291,25 @@ "## Target estimand\n", "Estimand type: ate\n", "### Estimand : 1\n", + "Estimand name: iv\n", + "Estimand expression:\n", + "Expectation(Derivative(y, Z0)/Derivative(v, Z0))\n", + "Estimand assumption 1, Exclusion: If we remove {Z0,Z1}→v, then ¬(Z0,Z1→y)\n", + "Estimand assumption 2, As-if-random: If U→→y then ¬(U →→Z0,Z1)\n", + "### Estimand : 2\n", "Estimand name: backdoor\n", "Estimand expression:\n", "d \n", - "──(Expectation(y|X0,X2,X3,X1,X4))\n", + "──(Expectation(y|X0,X1,X4,X3,X2))\n", "dv \n", - "Estimand assumption 1, Unconfoundedness: If U→v and U→y then P(y|v,X0,X2,X3,X1,X4,U) = P(y|v,X0,X2,X3,X1,X4)\n", - "### Estimand : 2\n", - "Estimand name: iv\n", - "Estimand expression:\n", - "Expectation(Derivative(y, Z1)/Derivative(v, Z1))\n", - "Estimand assumption 1, As-if-random: If U→→y then ¬(U →→Z1,Z0)\n", - "Estimand assumption 2, Exclusion: If we remove {Z1,Z0}→v, then ¬(Z1,Z0→y)\n", + "Estimand assumption 1, Unconfoundedness: If U→v and U→y then P(y|v,X0,X1,X4,X3,X2,U) = P(y|v,X0,X1,X4,X3,X2)\n", "\n", "## Realized estimand\n", - "b: y~v+X0+X2+X3+X1+X4\n", + "b: y~v+X0+X1+X4+X3+X2\n", "## Estimate\n", - "Value: 10.010658729918758\n", + "Value: 9.935979154269207\n", "\n", - "Causal Estimate is 10.010658729918758\n" - ] - }, - { - "name": "stderr", - "output_type": "stream", - "text": [ - "/home/amit/virtualenvs/python37/lib/python3.7/site-packages/sklearn/linear_model/logistic.py:432: FutureWarning: Default solver will be changed to 'lbfgs' in 0.22. Specify a solver to silence this warning.\n", - " FutureWarning)\n" + "Causal Estimate is 9.93597915427\n" ] } ], @@ -362,7 +339,7 @@ "output_type": "stream", "text": [ "INFO:dowhy.causal_estimator:INFO: Using Propensity Score Matching Estimator\n", - "INFO:dowhy.causal_estimator:b: y~v+X0+X2+X3+X1+X4\n" + "INFO:dowhy.causal_estimator:b: y~v+X0+X1+X4+X3+X2\n" ] }, { @@ -374,25 +351,25 @@ "## Target estimand\n", "Estimand type: ate\n", "### Estimand : 1\n", + "Estimand name: iv\n", + "Estimand expression:\n", + "Expectation(Derivative(y, Z0)/Derivative(v, Z0))\n", + "Estimand assumption 1, Exclusion: If we remove {Z0,Z1}→v, then ¬(Z0,Z1→y)\n", + "Estimand assumption 2, As-if-random: If U→→y then ¬(U →→Z0,Z1)\n", + "### Estimand : 2\n", "Estimand name: backdoor\n", "Estimand expression:\n", "d \n", - "──(Expectation(y|X0,X2,X3,X1,X4))\n", + "──(Expectation(y|X0,X1,X4,X3,X2))\n", "dv \n", - "Estimand assumption 1, Unconfoundedness: If U→v and U→y then P(y|v,X0,X2,X3,X1,X4,U) = P(y|v,X0,X2,X3,X1,X4)\n", - "### Estimand : 2\n", - "Estimand name: iv\n", - "Estimand expression:\n", - "Expectation(Derivative(y, Z1)/Derivative(v, Z1))\n", - "Estimand assumption 1, As-if-random: If U→→y then ¬(U →→Z1,Z0)\n", - "Estimand assumption 2, Exclusion: If we remove {Z1,Z0}→v, then ¬(Z1,Z0→y)\n", + "Estimand assumption 1, Unconfoundedness: If U→v and U→y then P(y|v,X0,X1,X4,X3,X2,U) = P(y|v,X0,X1,X4,X3,X2)\n", "\n", "## Realized estimand\n", - "b: y~v+X0+X2+X3+X1+X4\n", + "b: y~v+X0+X1+X4+X3+X2\n", "## Estimate\n", - "Value: 10.023841101622502\n", + "Value: 9.69335705963031\n", "\n", - "Causal Estimate is 10.023841101622502\n" + "Causal Estimate is 9.69335705963031\n" ] } ], @@ -425,7 +402,7 @@ "output_type": "stream", "text": [ "INFO:dowhy.causal_estimator:INFO: Using Propensity Score Weighting Estimator\n", - "INFO:dowhy.causal_estimator:b: y~v+X0+X2+X3+X1+X4\n" + "INFO:dowhy.causal_estimator:b: y~v+X0+X1+X4+X3+X2\n" ] }, { @@ -437,33 +414,25 @@ "## Target estimand\n", "Estimand type: ate\n", "### Estimand : 1\n", + "Estimand name: iv\n", + "Estimand expression:\n", + "Expectation(Derivative(y, Z0)/Derivative(v, Z0))\n", + "Estimand assumption 1, Exclusion: If we remove {Z0,Z1}→v, then ¬(Z0,Z1→y)\n", + "Estimand assumption 2, As-if-random: If U→→y then ¬(U →→Z0,Z1)\n", + "### Estimand : 2\n", "Estimand name: backdoor\n", "Estimand expression:\n", "d \n", - "──(Expectation(y|X0,X2,X3,X1,X4))\n", + "──(Expectation(y|X0,X1,X4,X3,X2))\n", "dv \n", - "Estimand assumption 1, Unconfoundedness: If U→v and U→y then P(y|v,X0,X2,X3,X1,X4,U) = P(y|v,X0,X2,X3,X1,X4)\n", - "### Estimand : 2\n", - "Estimand name: iv\n", - "Estimand expression:\n", - "Expectation(Derivative(y, Z1)/Derivative(v, Z1))\n", - "Estimand assumption 1, As-if-random: If U→→y then ¬(U →→Z1,Z0)\n", - "Estimand assumption 2, Exclusion: If we remove {Z1,Z0}→v, then ¬(Z1,Z0→y)\n", + "Estimand assumption 1, Unconfoundedness: If U→v and U→y then P(y|v,X0,X1,X4,X3,X2,U) = P(y|v,X0,X1,X4,X3,X2)\n", "\n", "## Realized estimand\n", - "b: y~v+X0+X2+X3+X1+X4\n", + "b: y~v+X0+X1+X4+X3+X2\n", "## Estimate\n", - "Value: 11.650976674743392\n", + "Value: 10.578537837065253\n", "\n", - "Causal Estimate is 11.650976674743392\n" - ] - }, - { - "name": "stderr", - "output_type": "stream", - "text": [ - "/home/amit/virtualenvs/python37/lib/python3.7/site-packages/sklearn/linear_model/logistic.py:432: FutureWarning: Default solver will be changed to 'lbfgs' in 0.22. Specify a solver to silence this warning.\n", - " FutureWarning)\n" + "Causal Estimate is 10.5785378371\n" ] } ], @@ -485,7 +454,7 @@ }, { "cell_type": "code", - "execution_count": 16, + "execution_count": 12, "metadata": {}, "outputs": [ { @@ -497,10 +466,10 @@ "Realized estimand type: ate\n", "Estimand expression:\n", " -1\n", - "Expectation(Derivative(y, Z1))⋅Expectation(Derivative(v, Z1)) \n", - "Estimand assumption 1, As-if-random: If U→→y then ¬(U →→Z1,Z0)\n", - "Estimand assumption 2, Exclusion: If we remove {Z1,Z0}→v, then ¬(Z1,Z0→y)\n", - "Estimand assumption 3, treatment_effect_homogeneity: Each unit's treatment v isaffected in the same way by common causes of v and y\n", + "Expectation(Derivative(y, Z0))⋅Expectation(Derivative(v, Z0)) \n", + "Estimand assumption 1, treatment_effect_homogeneity: Each unit's treatment v isaffected in the same way by common causes of v and y\n", + "Estimand assumption 2, Exclusion: If we remove {Z0,Z1}→v, then ¬(Z0,Z1→y)\n", + "Estimand assumption 3, As-if-random: If U→→y then ¬(U →→Z0,Z1)\n", "Estimand assumption 4, outcome_effect_homogeneity: Each unit's outcome y isaffected in the same way by common causes of v and y\n", "\n" ] @@ -514,34 +483,34 @@ "## Target estimand\n", "Estimand type: ate\n", "### Estimand : 1\n", + "Estimand name: iv\n", + "Estimand expression:\n", + "Expectation(Derivative(y, Z0)/Derivative(v, Z0))\n", + "Estimand assumption 1, Exclusion: If we remove {Z0,Z1}→v, then ¬(Z0,Z1→y)\n", + "Estimand assumption 2, As-if-random: If U→→y then ¬(U →→Z0,Z1)\n", + "### Estimand : 2\n", "Estimand name: backdoor\n", "Estimand expression:\n", "d \n", - "──(Expectation(y|X0,X2,X3,X1,X4))\n", + "──(Expectation(y|X0,X1,X4,X3,X2))\n", "dv \n", - "Estimand assumption 1, Unconfoundedness: If U→v and U→y then P(y|v,X0,X2,X3,X1,X4,U) = P(y|v,X0,X2,X3,X1,X4)\n", - "### Estimand : 2\n", - "Estimand name: iv\n", - "Estimand expression:\n", - "Expectation(Derivative(y, Z1)/Derivative(v, Z1))\n", - "Estimand assumption 1, As-if-random: If U→→y then ¬(U →→Z1,Z0)\n", - "Estimand assumption 2, Exclusion: If we remove {Z1,Z0}→v, then ¬(Z1,Z0→y)\n", + "Estimand assumption 1, Unconfoundedness: If U→v and U→y then P(y|v,X0,X1,X4,X3,X2,U) = P(y|v,X0,X1,X4,X3,X2)\n", "\n", "## Realized estimand\n", "Realized estimand: Wald Estimator\n", "Realized estimand type: ate\n", "Estimand expression:\n", " -1\n", - "Expectation(Derivative(y, Z1))⋅Expectation(Derivative(v, Z1)) \n", - "Estimand assumption 1, As-if-random: If U→→y then ¬(U →→Z1,Z0)\n", - "Estimand assumption 2, Exclusion: If we remove {Z1,Z0}→v, then ¬(Z1,Z0→y)\n", - "Estimand assumption 3, treatment_effect_homogeneity: Each unit's treatment v isaffected in the same way by common causes of v and y\n", + "Expectation(Derivative(y, Z0))⋅Expectation(Derivative(v, Z0)) \n", + "Estimand assumption 1, treatment_effect_homogeneity: Each unit's treatment v isaffected in the same way by common causes of v and y\n", + "Estimand assumption 2, Exclusion: If we remove {Z0,Z1}→v, then ¬(Z0,Z1→y)\n", + "Estimand assumption 3, As-if-random: If U→→y then ¬(U →→Z0,Z1)\n", "Estimand assumption 4, outcome_effect_homogeneity: Each unit's outcome y isaffected in the same way by common causes of v and y\n", "\n", "## Estimate\n", - "Value: 10.05993236042058\n", + "Value: 10.049189320441235\n", "\n", - "Causal Estimate is 10.05993236042058\n" + "Causal Estimate is 10.0491893204\n" ] } ], @@ -563,7 +532,7 @@ }, { "cell_type": "code", - "execution_count": 14, + "execution_count": 13, "metadata": {}, "outputs": [ { @@ -577,10 +546,10 @@ "Realized estimand type: ate\n", "Estimand expression:\n", " -1\n", - "Expectation(Derivative(y, Z1))⋅Expectation(Derivative(v, Z1)) \n", - "Estimand assumption 1, As-if-random: If U→→y then ¬(U →→Z1,Z0)\n", - "Estimand assumption 2, Exclusion: If we remove {Z1,Z0}→v, then ¬(Z1,Z0→y)\n", - "Estimand assumption 3, treatment_effect_homogeneity: Each unit's treatment local_treatment isaffected in the same way by common causes of local_treatment and local_outcome\n", + "Expectation(Derivative(y, Z0))⋅Expectation(Derivative(v, Z0)) \n", + "Estimand assumption 1, treatment_effect_homogeneity: Each unit's treatment local_treatment isaffected in the same way by common causes of local_treatment and local_outcome\n", + "Estimand assumption 2, Exclusion: If we remove {Z0,Z1}→v, then ¬(Z0,Z1→y)\n", + "Estimand assumption 3, As-if-random: If U→→y then ¬(U →→Z0,Z1)\n", "Estimand assumption 4, outcome_effect_homogeneity: Each unit's outcome local_outcome isaffected in the same way by common causes of local_treatment and local_outcome\n", "\n" ] @@ -594,34 +563,34 @@ "## Target estimand\n", "Estimand type: ate\n", "### Estimand : 1\n", + "Estimand name: iv\n", + "Estimand expression:\n", + "Expectation(Derivative(y, Z0)/Derivative(v, Z0))\n", + "Estimand assumption 1, Exclusion: If we remove {Z0,Z1}→v, then ¬(Z0,Z1→y)\n", + "Estimand assumption 2, As-if-random: If U→→y then ¬(U →→Z0,Z1)\n", + "### Estimand : 2\n", "Estimand name: backdoor\n", "Estimand expression:\n", "d \n", - "──(Expectation(y|X0,X2,X3,X1,X4))\n", + "──(Expectation(y|X0,X1,X4,X3,X2))\n", "dv \n", - "Estimand assumption 1, Unconfoundedness: If U→v and U→y then P(y|v,X0,X2,X3,X1,X4,U) = P(y|v,X0,X2,X3,X1,X4)\n", - "### Estimand : 2\n", - "Estimand name: iv\n", - "Estimand expression:\n", - "Expectation(Derivative(y, Z1)/Derivative(v, Z1))\n", - "Estimand assumption 1, As-if-random: If U→→y then ¬(U →→Z1,Z0)\n", - "Estimand assumption 2, Exclusion: If we remove {Z1,Z0}→v, then ¬(Z1,Z0→y)\n", + "Estimand assumption 1, Unconfoundedness: If U→v and U→y then P(y|v,X0,X1,X4,X3,X2,U) = P(y|v,X0,X1,X4,X3,X2)\n", "\n", "## Realized estimand\n", "Realized estimand: Wald Estimator\n", "Realized estimand type: ate\n", "Estimand expression:\n", " -1\n", - "Expectation(Derivative(y, Z1))⋅Expectation(Derivative(v, Z1)) \n", - "Estimand assumption 1, As-if-random: If U→→y then ¬(U →→Z1,Z0)\n", - "Estimand assumption 2, Exclusion: If we remove {Z1,Z0}→v, then ¬(Z1,Z0→y)\n", - "Estimand assumption 3, treatment_effect_homogeneity: Each unit's treatment local_treatment isaffected in the same way by common causes of local_treatment and local_outcome\n", + "Expectation(Derivative(y, Z0))⋅Expectation(Derivative(v, Z0)) \n", + "Estimand assumption 1, treatment_effect_homogeneity: Each unit's treatment local_treatment isaffected in the same way by common causes of local_treatment and local_outcome\n", + "Estimand assumption 2, Exclusion: If we remove {Z0,Z1}→v, then ¬(Z0,Z1→y)\n", + "Estimand assumption 3, As-if-random: If U→→y then ¬(U →→Z0,Z1)\n", "Estimand assumption 4, outcome_effect_homogeneity: Each unit's outcome local_outcome isaffected in the same way by common causes of local_treatment and local_outcome\n", "\n", "## Estimate\n", - "Value: 14.837556010815915\n", + "Value: 1.0\n", "\n", - "Causal Estimate is 14.837556010815915\n" + "Causal Estimate is 1.0\n" ] } ], @@ -652,7 +621,7 @@ "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", - "version": "3.7.4" + "version": "3.5.2" } }, "nbformat": 4, diff --git a/dowhy/causal_estimators/instrumental_variable_estimator.py b/dowhy/causal_estimators/instrumental_variable_estimator.py index 288e084df..a6abfe98b 100644 --- a/dowhy/causal_estimators/instrumental_variable_estimator.py +++ b/dowhy/causal_estimators/instrumental_variable_estimator.py @@ -47,7 +47,7 @@ class InstrumentalVariableEstimator(CausalEstimator): else: # Obtain estimate by 2SLS estimator: Cov(y,z) / Cov(x,z) num_yz = np.cov(self._outcome, instrument)[0, 1] - deno_xz = np.cov(self._outcome, instrument)[0, 1] + deno_xz = np.cov(self._treatment, instrument)[0, 1] iv_est = num_yz / deno_xz estimate = CausalEstimate(estimate=iv_est,