CircleCI build openturns 18113

openturns/master/_downloads/3a76cbe9f372eb9e6634c1a8f6ee07ed/plot_sensitivity_rank_sobol.ipynb

@@ -0,0 +1,176 @@
+{
+  "cells": [
+    {
+      "cell_type": "markdown",
+      "metadata": {},
+      "source": [
+        "\n# Sobol' sensitivity indices using rank-based algorithm\n"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "collapsed": false
+      },
+      "outputs": [],
+      "source": [
+        "import openturns as ot\nimport openturns.experimental as otexp\nimport openturns.viewer as viewer\nfrom openturns.usecases import ishigami_function"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {},
+      "source": [
+        "In this example we are going to compute global sensitivity indices in a data driven framework.\nFor that purpose, we are using the rank-based algorithm which provides the first order Sobol' indices.\nWe present the method on the `Ishigami function<use-case-ishigami>`.\n\n"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "collapsed": false
+      },
+      "outputs": [],
+      "source": [
+        "ot.Log.Show(ot.Log.NONE)"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {},
+      "source": [
+        "## Definition of the model\n\nWe load the model from the usecases module.\n\n"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "collapsed": false
+      },
+      "outputs": [],
+      "source": [
+        "im = ishigami_function.IshigamiModel()"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {},
+      "source": [
+        "We generate an input sample of size 500 (and dimension 3).\n\n"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "collapsed": false
+      },
+      "outputs": [],
+      "source": [
+        "size = 500\not.RandomGenerator.SetSeed(10)\nX = im.distributionX.getSample(size)"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {},
+      "source": [
+        "We compute the output by applying the Ishigami model to the input sample.\n\n"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "collapsed": false
+      },
+      "outputs": [],
+      "source": [
+        "Y = im.model(X)"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {},
+      "source": [
+        "Estimate Sobol' first order indices via rank-based algorithm.\n\n"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "collapsed": false
+      },
+      "outputs": [],
+      "source": [
+        "mySobol = otexp.RankSobolSensitivityAlgorithm(X, Y)\nindices = mySobol.getFirstOrderIndices()\nprint('First order indices:', indices)"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {},
+      "source": [
+        "Draw Sobol' indices.\n\nOptions for confidence interval estimation can be defined.\n\n"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "collapsed": false
+      },
+      "outputs": [],
+      "source": [
+        "ot.ResourceMap.SetAsUnsignedInteger('SobolIndicesAlgorithm-DefaultBootstrapSize', 200)\not.ResourceMap.SetAsScalar('RankSobolSensitivityAlgorithm-DefaultBootstrapSampleRatio', 0.85)"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {},
+      "source": [
+        "if SobolIndicesAlgorithm-DefaultBootstrapSize > 1, the draw() method will call bootstrap method to compute the confidence intervals.\n\n"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "collapsed": false
+      },
+      "outputs": [],
+      "source": [
+        "graph = mySobol.draw()\ngraph.setTitle(\"Sobol' indices\")\nview = viewer.View(graph)\n\nprint('confidence intervals:', mySobol.getFirstOrderIndicesInterval())"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {},
+      "source": [
+        "It is possible that Sobol' indices estimates take negative values, that is inconsistent with the theory. Therefore, a larger number of samples is required to get consistent indices.\n\n"
+      ]
+    }
+  ],
+  "metadata": {
+    "kernelspec": {
+      "display_name": "Python 3",
+      "language": "python",
+      "name": "python3"
+    },
+    "language_info": {
+      "codemirror_mode": {
+        "name": "ipython",
+        "version": 3
+      },
+      "file_extension": ".py",
+      "mimetype": "text/x-python",
+      "name": "python",
+      "nbconvert_exporter": "python",
+      "pygments_lexer": "ipython3",
+      "version": "3.11.8"
+    }
+  },
+  "nbformat": 4,
+  "nbformat_minor": 0
+}

openturns/master/_downloads/5c34535a049aaa69f75a4fb84a6df6c4/plot_flood_model.ipynb

@@ -151,7 +151,7 @@
       },
       "outputs": [],
       "source": [
-        "optimAlgo = ot.Cobyla()\noptimAlgo.setMaximumCallsNumber(1000)\noptimAlgo.setMaximumAbsoluteError(1.0e-10)\noptimAlgo.setMaximumRelativeError(1.0e-10)\noptimAlgo.setMaximumResidualError(1.0e-10)\noptimAlgo.setMaximumConstraintError(1.0e-10)"
+        "optimAlgo = ot.Cobyla()\noptimAlgo.setMaximumCallsNumber(1000)\noptimAlgo.setMaximumAbsoluteError(1.0e-8)\noptimAlgo.setMaximumRelativeError(1.0e-10)\noptimAlgo.setMaximumResidualError(1.0e-10)\noptimAlgo.setMaximumConstraintError(1.0e-10)"
       ]
     },
     {

openturns/master/_downloads/8b7669cacbda382af56e0d4f8e63131b/plot_flood_model.py

@@ -64,7 +64,7 @@
 # %%
 optimAlgo = ot.Cobyla()
 optimAlgo.setMaximumCallsNumber(1000)
-optimAlgo.setMaximumAbsoluteError(1.0e-10)
+optimAlgo.setMaximumAbsoluteError(1.0e-8)
 optimAlgo.setMaximumRelativeError(1.0e-10)
 optimAlgo.setMaximumResidualError(1.0e-10)
 optimAlgo.setMaximumConstraintError(1.0e-10)

openturns/master/_downloads/b77fabceeed0488793aa1923f45f794f/plot_sensitivity_wingweight.py

@@ -14,6 +14,7 @@
 # - Spearman coefficients
 # - Taylor expansion importance factors
 # - Sobol' indices
+# - Rank-based estimation of Sobol' indices
 # - HSIC : Hilbert-Schmidt Independence Criterion
 #
 # We present the methods on the :ref:`WingWeight function<use-case-wingweight>` and use the same notations.
@@ -26,6 +27,7 @@
 #
 #
 import openturns as ot
+import openturns.experimental as otexp
 import openturns.viewer as otv
 from openturns.usecases.wingweight_function import WingWeightModel
 from matplotlib import pylab as plt
@@ -366,6 +368,20 @@
 graph.setTitle("Sobol indices by Polynomial Chaos Expansion - wing weight")
 view = otv.View(graph)
 
+
+# %%
+# Furthermore, first order Sobol' indices can also been estimated in a data-driven way using a rank-based sensitivity algorithm.
+# In such a way, the estimation of sensitivity indices does not involve any surrogate model.
+sizeRankSobol = 800
+inputDesignRankSobol = m.distributionX.getSample(sizeRankSobol)
+outputDesignankSobol = m.model(inputDesignRankSobol)
+myRankSobol = otexp.RankSobolSensitivityAlgorithm(inputDesignRankSobol, outputDesignankSobol)
+indicesrankSobol = myRankSobol.getFirstOrderIndices()
+print('First order indices:', indicesrankSobol)
+graph = myRankSobol.draw()
+graph.setTitle("Sobol indices by rank-based estimation - wing weight")
+view = otv.View(graph)
+
 # %%
 #
 # The Sobol' indices confirm the previous analyses, in terms of ranking of the most influent variables.

openturns/master/_downloads/ca1cafbe970135714f4ef620e84fc370/plot_sensitivity_wingweight.ipynb

@@ -11,7 +11,7 @@
       "cell_type": "markdown",
       "metadata": {},
       "source": [
-        "This example is a brief overview of the use of the most usual sensitivity analysis techniques and how to call them:\n\n- PCC: Partial Correlation Coefficients\n- PRCC: Partial Rank Correlation Coefficients\n- SRC: Standard Regression Coefficients\n- SRRC: Standard Rank Regression Coefficients\n- Pearson coefficients\n- Spearman coefficients\n- Taylor expansion importance factors\n- Sobol' indices\n- HSIC : Hilbert-Schmidt Independence Criterion\n\nWe present the methods on the `WingWeight function<use-case-wingweight>` and use the same notations.\n\n"
+        "This example is a brief overview of the use of the most usual sensitivity analysis techniques and how to call them:\n\n- PCC: Partial Correlation Coefficients\n- PRCC: Partial Rank Correlation Coefficients\n- SRC: Standard Regression Coefficients\n- SRRC: Standard Rank Regression Coefficients\n- Pearson coefficients\n- Spearman coefficients\n- Taylor expansion importance factors\n- Sobol' indices\n- Rank-based estimation of Sobol' indices\n- HSIC : Hilbert-Schmidt Independence Criterion\n\nWe present the methods on the `WingWeight function<use-case-wingweight>` and use the same notations.\n\n"
       ]
     },
     {
@@ -29,7 +29,7 @@
       },
       "outputs": [],
       "source": [
-        "import openturns as ot\nimport openturns.viewer as otv\nfrom openturns.usecases.wingweight_function import WingWeightModel\nfrom matplotlib import pylab as plt\nimport numpy as np\n\not.Log.Show(ot.Log.NONE)\nm = WingWeightModel()"
+        "import openturns as ot\nimport openturns.experimental as otexp\nimport openturns.viewer as otv\nfrom openturns.usecases.wingweight_function import WingWeightModel\nfrom matplotlib import pylab as plt\nimport numpy as np\n\not.Log.Show(ot.Log.NONE)\nm = WingWeightModel()"
       ]
     },
     {
@@ -594,6 +594,24 @@
         "firstOrder = [sensitivityAnalysis.getSobolIndex(i) for i in range(m.dim)]\ntotalOrder = [sensitivityAnalysis.getSobolTotalIndex(i) for i in range(m.dim)]\ngraph = ot.SobolIndicesAlgorithm.DrawSobolIndices(inputNames, firstOrder, totalOrder)\ngraph.setTitle(\"Sobol indices by Polynomial Chaos Expansion - wing weight\")\nview = otv.View(graph)"
       ]
     },
+    {
+      "cell_type": "markdown",
+      "metadata": {},
+      "source": [
+        "Furthermore, first order Sobol' indices can also been estimated in a data-driven way using a rank-based sensitivity algorithm.\nIn such a way, the estimation of sensitivity indices does not involve any surrogate model.\n\n"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "collapsed": false
+      },
+      "outputs": [],
+      "source": [
+        "sizeRankSobol = 800\ninputDesignRankSobol = m.distributionX.getSample(sizeRankSobol)\noutputDesignankSobol = m.model(inputDesignRankSobol)\nmyRankSobol = otexp.RankSobolSensitivityAlgorithm(inputDesignRankSobol, outputDesignankSobol)\nindicesrankSobol = myRankSobol.getFirstOrderIndices()\nprint('First order indices:', indicesrankSobol)\ngraph = myRankSobol.draw()\ngraph.setTitle(\"Sobol indices by rank-based estimation - wing weight\")\nview = otv.View(graph)"
+      ]
+    },
     {
       "cell_type": "markdown",
       "metadata": {},

openturns/master/_downloads/fb2afce0346a87308e38e6b4d9b7197d/plot_sensitivity_rank_sobol.py

@@ -0,0 +1,58 @@
+"""
+Sobol' sensitivity indices using rank-based algorithm
+=====================================================
+"""
+
+import openturns as ot
+import openturns.experimental as otexp
+import openturns.viewer as viewer
+from openturns.usecases import ishigami_function
+
+# %%
+# In this example we are going to compute global sensitivity indices in a data driven framework.
+# For that purpose, we are using the rank-based algorithm which provides the first order Sobol' indices.
+# We present the method on the :ref:`Ishigami function<use-case-ishigami>`.
+
+ot.Log.Show(ot.Log.NONE)
+
+# %%
+# Definition of the model
+# -----------------------
+#
+# We load the model from the usecases module.
+im = ishigami_function.IshigamiModel()
+
+# %%
+# We generate an input sample of size 500 (and dimension 3).
+size = 500
+ot.RandomGenerator.SetSeed(10)
+X = im.distributionX.getSample(size)
+
+# %%
+# We compute the output by applying the Ishigami model to the input sample.
+Y = im.model(X)
+
+# %%
+# Estimate Sobol' first order indices via rank-based algorithm.
+mySobol = otexp.RankSobolSensitivityAlgorithm(X, Y)
+indices = mySobol.getFirstOrderIndices()
+print('First order indices:', indices)
+
+# %%
+# Draw Sobol' indices.
+#
+# Options for confidence interval estimation can be defined.
+ot.ResourceMap.SetAsUnsignedInteger('SobolIndicesAlgorithm-DefaultBootstrapSize', 200)
+ot.ResourceMap.SetAsScalar('RankSobolSensitivityAlgorithm-DefaultBootstrapSampleRatio', 0.85)
+
+# %%
+# if SobolIndicesAlgorithm-DefaultBootstrapSize > 1, the draw() method will call bootstrap method to compute the confidence intervals.
+
+graph = mySobol.draw()
+graph.setTitle("Sobol' indices")
+view = viewer.View(graph)
+
+print('confidence intervals:', mySobol.getFirstOrderIndicesInterval())
+
+# %%
+# It is possible that Sobol' indices estimates take negative values, that is inconsistent with the theory. Therefore, a larger number of samples is required to get consistent indices.

openturns/master/_sources/auto_calibration/bayesian_calibration/plot_gibbs.rst.txt

@@ -283,7 +283,7 @@ Let us plot the posterior density.
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** (0 minutes 12.024 seconds)
+   **Total running time of the script:** (0 minutes 13.233 seconds)
 
 
 .. _sphx_glr_download_auto_calibration_bayesian_calibration_plot_gibbs.py:

openturns/master/_sources/auto_calibration/bayesian_calibration/plot_rwmh_python_distribution.rst.txt

@@ -425,7 +425,7 @@ Plot posterior marginal plots only as prior cannot be drawn meaningfully.
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** (0 minutes 5.445 seconds)
+   **Total running time of the script:** (0 minutes 5.731 seconds)
 
 
 .. _sphx_glr_download_auto_calibration_bayesian_calibration_plot_rwmh_python_distribution.py:

openturns/master/_sources/auto_calibration/bayesian_calibration/sg_execution_times.rst.txt

@@ -6,7 +6,7 @@
 
 Computation times
 =================
-**00:22.266** total execution time for 7 files **from auto_calibration/bayesian_calibration**:
+**00:24.024** total execution time for 7 files **from auto_calibration/bayesian_calibration**:
 
 .. container::
 
@@ -33,23 +33,23 @@ Computation times
      - Time
      - Mem (MB)
    * - :ref:`sphx_glr_auto_calibration_bayesian_calibration_plot_gibbs.py` (``plot_gibbs.py``)
-     - 00:12.024
+     - 00:13.233
      - 0.0
    * - :ref:`sphx_glr_auto_calibration_bayesian_calibration_plot_rwmh_python_distribution.py` (``plot_rwmh_python_distribution.py``)
-     - 00:05.445
+     - 00:05.731
      - 0.0
    * - :ref:`sphx_glr_auto_calibration_bayesian_calibration_plot_bayesian_calibration_flooding.py` (``plot_bayesian_calibration_flooding.py``)
-     - 00:01.332
+     - 00:01.408
      - 0.0
    * - :ref:`sphx_glr_auto_calibration_bayesian_calibration_plot_bayesian_calibration.py` (``plot_bayesian_calibration.py``)
-     - 00:01.217
+     - 00:01.209
      - 0.0
    * - :ref:`sphx_glr_auto_calibration_bayesian_calibration_plot_gibbs_simus.py` (``plot_gibbs_simus.py``)
-     - 00:01.097
+     - 00:01.188
      - 0.0
    * - :ref:`sphx_glr_auto_calibration_bayesian_calibration_plot_ackley_distribution.py` (``plot_ackley_distribution.py``)
-     - 00:00.882
+     - 00:00.961
      - 0.0
    * - :ref:`sphx_glr_auto_calibration_bayesian_calibration_plot_imh_python_distribution.py` (``plot_imh_python_distribution.py``)
-     - 00:00.269
+     - 00:00.294
      - 0.0