|
| 1 | +{ |
| 2 | + "cells": [ |
| 3 | + { |
| 4 | + "cell_type": "markdown", |
| 5 | + "metadata": {}, |
| 6 | + "source": [ |
| 7 | + "# SymbolicBayesNet" |
| 8 | + ] |
| 9 | + }, |
| 10 | + { |
| 11 | + "cell_type": "markdown", |
| 12 | + "metadata": {}, |
| 13 | + "source": [ |
| 14 | + "A `SymbolicBayesNet` is a directed acyclic graph (DAG) composed of `SymbolicConditional` objects. It represents the structure of a factorized probability distribution P(X) = Π P(Xi | Parents(Xi)) purely in terms of variable connectivity.\n", |
| 15 | + "\n", |
| 16 | + "It is typically the result of running sequential variable elimination on a `SymbolicFactorGraph`." |
| 17 | + ] |
| 18 | + }, |
| 19 | + { |
| 20 | + "cell_type": "markdown", |
| 21 | + "metadata": {}, |
| 22 | + "source": [ |
| 23 | + "<a href=\"https://colab.research.google.com/github/borglab/gtsam/blob/develop/gtsam/symbolic/doc/SymbolicBayesNet.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>" |
| 24 | + ] |
| 25 | + }, |
| 26 | + { |
| 27 | + "cell_type": "code", |
| 28 | + "execution_count": 13, |
| 29 | + "metadata": { |
| 30 | + "tags": [ |
| 31 | + "remove-cell" |
| 32 | + ] |
| 33 | + }, |
| 34 | + "outputs": [ |
| 35 | + { |
| 36 | + "name": "stdout", |
| 37 | + "output_type": "stream", |
| 38 | + "text": [ |
| 39 | + "Note: you may need to restart the kernel to use updated packages.\n" |
| 40 | + ] |
| 41 | + } |
| 42 | + ], |
| 43 | + "source": [ |
| 44 | + "%pip install --quiet gtsam-develop" |
| 45 | + ] |
| 46 | + }, |
| 47 | + { |
| 48 | + "cell_type": "code", |
| 49 | + "execution_count": 14, |
| 50 | + "metadata": {}, |
| 51 | + "outputs": [], |
| 52 | + "source": [ |
| 53 | + "from gtsam import SymbolicConditional, SymbolicFactorGraph, Ordering\n", |
| 54 | + "from gtsam.symbol_shorthand import X, L\n", |
| 55 | + "import graphviz" |
| 56 | + ] |
| 57 | + }, |
| 58 | + { |
| 59 | + "cell_type": "markdown", |
| 60 | + "metadata": {}, |
| 61 | + "source": [ |
| 62 | + "## Creating a SymbolicBayesNet\n", |
| 63 | + "\n", |
| 64 | + "SymbolicBayesNets are usually created by eliminating a [SymbolicFactorGraph](SymbolicFactorGraph.ipynb). But you can also build them directly:" |
| 65 | + ] |
| 66 | + }, |
| 67 | + { |
| 68 | + "cell_type": "code", |
| 69 | + "execution_count": 15, |
| 70 | + "metadata": {}, |
| 71 | + "outputs": [ |
| 72 | + { |
| 73 | + "name": "stdout", |
| 74 | + "output_type": "stream", |
| 75 | + "text": [ |
| 76 | + "Directly Built Symbolic Bayes Net:\n", |
| 77 | + " \n", |
| 78 | + "size: 5\n", |
| 79 | + "conditional 0: P( l1 | x0)\n", |
| 80 | + "conditional 1: P( x0 | x1)\n", |
| 81 | + "conditional 2: P( l2 | x1)\n", |
| 82 | + "conditional 3: P( x1 | x2)\n", |
| 83 | + "conditional 4: P( x2)\n" |
| 84 | + ] |
| 85 | + } |
| 86 | + ], |
| 87 | + "source": [ |
| 88 | + "from gtsam import SymbolicBayesNet\n", |
| 89 | + "\n", |
| 90 | + "# Create a new Bayes Net\n", |
| 91 | + "symbolic_bayes_net = SymbolicBayesNet()\n", |
| 92 | + "\n", |
| 93 | + "# Add conditionals directly\n", |
| 94 | + "symbolic_bayes_net.push_back(SymbolicConditional(L(1), X(0))) # P(l1 | x0)\n", |
| 95 | + "symbolic_bayes_net.push_back(SymbolicConditional(X(0), X(1))) # P(x0 | x1)\n", |
| 96 | + "symbolic_bayes_net.push_back(SymbolicConditional(L(2), X(1))) # P(l2 | x1)\n", |
| 97 | + "symbolic_bayes_net.push_back(SymbolicConditional(X(1), X(2))) # P(x1 | x2)\n", |
| 98 | + "symbolic_bayes_net.push_back(SymbolicConditional(X(2))) # P(x2)\n", |
| 99 | + "\n", |
| 100 | + "symbolic_bayes_net.print(\"Directly Built Symbolic Bayes Net:\\n\")" |
| 101 | + ] |
| 102 | + }, |
| 103 | + { |
| 104 | + "cell_type": "markdown", |
| 105 | + "metadata": {}, |
| 106 | + "source": [ |
| 107 | + "## Accessing Conditionals and Visualization" |
| 108 | + ] |
| 109 | + }, |
| 110 | + { |
| 111 | + "cell_type": "code", |
| 112 | + "execution_count": 16, |
| 113 | + "metadata": {}, |
| 114 | + "outputs": [ |
| 115 | + { |
| 116 | + "name": "stdout", |
| 117 | + "output_type": "stream", |
| 118 | + "text": [ |
| 119 | + "Conditional at index 1: P( x0 | x1)\n" |
| 120 | + ] |
| 121 | + }, |
| 122 | + { |
| 123 | + "data": { |
| 124 | + "image/svg+xml": [ |
| 125 | + "<?xml version=\"1.0\" encoding=\"UTF-8\" standalone=\"no\"?>\n", |
| 126 | + "<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\n", |
| 127 | + " \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\n", |
| 128 | + "<!-- Generated by graphviz version 12.0.0 (0)\n", |
| 129 | + " -->\n", |
| 130 | + "<!-- Pages: 1 -->\n", |
| 131 | + "<svg width=\"134pt\" height=\"260pt\"\n", |
| 132 | + " viewBox=\"0.00 0.00 134.00 260.00\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n", |
| 133 | + "<g id=\"graph0\" class=\"graph\" transform=\"scale(1 1) rotate(0) translate(4 256)\">\n", |
| 134 | + "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-256 130,-256 130,4 -4,4\"/>\n", |
| 135 | + "<!-- var7782220156096217089 -->\n", |
| 136 | + "<g id=\"node1\" class=\"node\">\n", |
| 137 | + "<title>var7782220156096217089</title>\n", |
| 138 | + "<ellipse fill=\"none\" stroke=\"black\" cx=\"99\" cy=\"-18\" rx=\"27\" ry=\"18\"/>\n", |
| 139 | + "<text text-anchor=\"middle\" x=\"99\" y=\"-12.95\" font-family=\"Times,serif\" font-size=\"14.00\">l1</text>\n", |
| 140 | + "</g>\n", |
| 141 | + "<!-- var7782220156096217090 -->\n", |
| 142 | + "<g id=\"node2\" class=\"node\">\n", |
| 143 | + "<title>var7782220156096217090</title>\n", |
| 144 | + "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-90\" rx=\"27\" ry=\"18\"/>\n", |
| 145 | + "<text text-anchor=\"middle\" x=\"27\" y=\"-84.95\" font-family=\"Times,serif\" font-size=\"14.00\">l2</text>\n", |
| 146 | + "</g>\n", |
| 147 | + "<!-- var8646911284551352320 -->\n", |
| 148 | + "<g id=\"node3\" class=\"node\">\n", |
| 149 | + "<title>var8646911284551352320</title>\n", |
| 150 | + "<ellipse fill=\"none\" stroke=\"black\" cx=\"99\" cy=\"-90\" rx=\"27\" ry=\"18\"/>\n", |
| 151 | + "<text text-anchor=\"middle\" x=\"99\" y=\"-84.95\" font-family=\"Times,serif\" font-size=\"14.00\">x0</text>\n", |
| 152 | + "</g>\n", |
| 153 | + "<!-- var8646911284551352320->var7782220156096217089 -->\n", |
| 154 | + "<g id=\"edge4\" class=\"edge\">\n", |
| 155 | + "<title>var8646911284551352320->var7782220156096217089</title>\n", |
| 156 | + "<path fill=\"none\" stroke=\"black\" d=\"M99,-71.7C99,-64.41 99,-55.73 99,-47.54\"/>\n", |
| 157 | + "<polygon fill=\"black\" stroke=\"black\" points=\"102.5,-47.62 99,-37.62 95.5,-47.62 102.5,-47.62\"/>\n", |
| 158 | + "</g>\n", |
| 159 | + "<!-- var8646911284551352321 -->\n", |
| 160 | + "<g id=\"node4\" class=\"node\">\n", |
| 161 | + "<title>var8646911284551352321</title>\n", |
| 162 | + "<ellipse fill=\"none\" stroke=\"black\" cx=\"63\" cy=\"-162\" rx=\"27\" ry=\"18\"/>\n", |
| 163 | + "<text text-anchor=\"middle\" x=\"63\" y=\"-156.95\" font-family=\"Times,serif\" font-size=\"14.00\">x1</text>\n", |
| 164 | + "</g>\n", |
| 165 | + "<!-- var8646911284551352321->var7782220156096217090 -->\n", |
| 166 | + "<g id=\"edge2\" class=\"edge\">\n", |
| 167 | + "<title>var8646911284551352321->var7782220156096217090</title>\n", |
| 168 | + "<path fill=\"none\" stroke=\"black\" d=\"M54.65,-144.76C50.42,-136.55 45.19,-126.37 40.42,-117.09\"/>\n", |
| 169 | + "<polygon fill=\"black\" stroke=\"black\" points=\"43.68,-115.79 36,-108.49 37.46,-118.99 43.68,-115.79\"/>\n", |
| 170 | + "</g>\n", |
| 171 | + "<!-- var8646911284551352321->var8646911284551352320 -->\n", |
| 172 | + "<g id=\"edge3\" class=\"edge\">\n", |
| 173 | + "<title>var8646911284551352321->var8646911284551352320</title>\n", |
| 174 | + "<path fill=\"none\" stroke=\"black\" d=\"M71.35,-144.76C75.58,-136.55 80.81,-126.37 85.58,-117.09\"/>\n", |
| 175 | + "<polygon fill=\"black\" stroke=\"black\" points=\"88.54,-118.99 90,-108.49 82.32,-115.79 88.54,-118.99\"/>\n", |
| 176 | + "</g>\n", |
| 177 | + "<!-- var8646911284551352322 -->\n", |
| 178 | + "<g id=\"node5\" class=\"node\">\n", |
| 179 | + "<title>var8646911284551352322</title>\n", |
| 180 | + "<ellipse fill=\"none\" stroke=\"black\" cx=\"63\" cy=\"-234\" rx=\"27\" ry=\"18\"/>\n", |
| 181 | + "<text text-anchor=\"middle\" x=\"63\" y=\"-228.95\" font-family=\"Times,serif\" font-size=\"14.00\">x2</text>\n", |
| 182 | + "</g>\n", |
| 183 | + "<!-- var8646911284551352322->var8646911284551352321 -->\n", |
| 184 | + "<g id=\"edge1\" class=\"edge\">\n", |
| 185 | + "<title>var8646911284551352322->var8646911284551352321</title>\n", |
| 186 | + "<path fill=\"none\" stroke=\"black\" d=\"M63,-215.7C63,-208.41 63,-199.73 63,-191.54\"/>\n", |
| 187 | + "<polygon fill=\"black\" stroke=\"black\" points=\"66.5,-191.62 63,-181.62 59.5,-191.62 66.5,-191.62\"/>\n", |
| 188 | + "</g>\n", |
| 189 | + "</g>\n", |
| 190 | + "</svg>\n" |
| 191 | + ], |
| 192 | + "text/plain": [ |
| 193 | + "<graphviz.sources.Source at 0x10c18fda0>" |
| 194 | + ] |
| 195 | + }, |
| 196 | + "metadata": {}, |
| 197 | + "output_type": "display_data" |
| 198 | + } |
| 199 | + ], |
| 200 | + "source": [ |
| 201 | + "# Access a conditional by index\n", |
| 202 | + "conditional_1 = bayes_net.at(1) # P(x0 | l1)\n", |
| 203 | + "conditional_1.print(\"Conditional at index 1: \")\n", |
| 204 | + "\n", |
| 205 | + "# Visualize the Bayes Net structure\n", |
| 206 | + "display(graphviz.Source(bayes_net.dot()))" |
| 207 | + ] |
| 208 | + } |
| 209 | + ], |
| 210 | + "metadata": { |
| 211 | + "kernelspec": { |
| 212 | + "display_name": "py312", |
| 213 | + "language": "python", |
| 214 | + "name": "python3" |
| 215 | + }, |
| 216 | + "language_info": { |
| 217 | + "codemirror_mode": { |
| 218 | + "name": "ipython", |
| 219 | + "version": 3 |
| 220 | + }, |
| 221 | + "file_extension": ".py", |
| 222 | + "mimetype": "text/x-python", |
| 223 | + "name": "python", |
| 224 | + "nbconvert_exporter": "python", |
| 225 | + "pygments_lexer": "ipython3", |
| 226 | + "version": "3.12.6" |
| 227 | + } |
| 228 | + }, |
| 229 | + "nbformat": 4, |
| 230 | + "nbformat_minor": 4 |
| 231 | +} |
0 commit comments