rmontanana/BayesNet
1
0
Fork 0
Code Issues 2 Pull Requests 1 Actions Packages Projects Releases 6 Wiki Activity

Add hyperparameter convergence_best

Browse Source 
move test libraries to test folder
This commit is contained in:
Ricardo Montañana Gómez
2024-04-30 00:52:09 +02:00
parent f014928411
commit ae469b8146
721 changed files with 206095 additions and 2496 deletions
Download Patch File Download Diff File Expand all files Collapse all files

48
html/bayesnet/classifiers/TAN.cc.gcov.html
View File

@@ -37,7 +37,7 @@
</tr>
<tr>
<td class="headerItem">Test Date:</td>
<td class="headerValue">2024-04-21 17:30:26</td>
<td class="headerValue">2024-04-29 20:48:03</td>
<td></td>
<td class="headerItem">Functions:</td>
<td class="headerCovTableEntryHi">100.0&nbsp;%</td>
@@ -69,40 +69,40 @@
<span id="L7"><span class="lineNum"> 7</span> : #include &quot;TAN.h&quot;</span>
<span id="L8"><span class="lineNum"> 8</span> : </span>
<span id="L9"><span class="lineNum"> 9</span> : namespace bayesnet {</span>
<span id="L10"><span class="lineNum"> 10</span> <span class="tlaGNC tlaBgGNC"> 47 : TAN::TAN() : Classifier(Network()) {}</span></span>
<span id="L10"><span class="lineNum"> 10</span> <span class="tlaGNC tlaBgGNC"> 517 : TAN::TAN() : Classifier(Network()) {}</span></span>
<span id="L11"><span class="lineNum"> 11</span> : </span>
<span id="L12"><span class="lineNum"> 12</span> <span class="tlaGNC"> 13 : void TAN::buildModel(const torch::Tensor&amp; weights)</span></span>
<span id="L12"><span class="lineNum"> 12</span> <span class="tlaGNC"> 143 : void TAN::buildModel(const torch::Tensor&amp; weights)</span></span>
<span id="L13"><span class="lineNum"> 13</span> : {</span>
<span id="L14"><span class="lineNum"> 14</span> : // 0. Add all nodes to the model</span>
<span id="L15"><span class="lineNum"> 15</span> <span class="tlaGNC"> 13 : addNodes();</span></span>
<span id="L15"><span class="lineNum"> 15</span> <span class="tlaGNC"> 143 : addNodes();</span></span>
<span id="L16"><span class="lineNum"> 16</span> : // 1. Compute mutual information between each feature and the class and set the root node</span>
<span id="L17"><span class="lineNum"> 17</span> : // as the highest mutual information with the class</span>
<span id="L18"><span class="lineNum"> 18</span> <span class="tlaGNC"> 13 : auto mi = std::vector &lt;std::pair&lt;int, float &gt;&gt;();</span></span>
<span id="L19"><span class="lineNum"> 19</span> <span class="tlaGNC"> 39 : torch::Tensor class_dataset = dataset.index({ -1, &quot;...&quot; });</span></span>
<span id="L20"><span class="lineNum"> 20</span> <span class="tlaGNC"> 89 : for (int i = 0; i &lt; static_cast&lt;int&gt;(features.size()); ++i) {</span></span>
<span id="L21"><span class="lineNum"> 21</span> <span class="tlaGNC"> 228 : torch::Tensor feature_dataset = dataset.index({ i, &quot;...&quot; });</span></span>
<span id="L22"><span class="lineNum"> 22</span> <span class="tlaGNC"> 76 : auto mi_value = metrics.mutualInformation(class_dataset, feature_dataset, weights);</span></span>
<span id="L23"><span class="lineNum"> 23</span> <span class="tlaGNC"> 76 : mi.push_back({ i, mi_value });</span></span>
<span id="L24"><span class="lineNum"> 24</span> <span class="tlaGNC"> 76 : }</span></span>
<span id="L25"><span class="lineNum"> 25</span> <span class="tlaGNC"> 175 : sort(mi.begin(), mi.end(), [](const auto&amp; left, const auto&amp; right) {return left.second &lt; right.second;});</span></span>
<span id="L26"><span class="lineNum"> 26</span> <span class="tlaGNC"> 13 : auto root = mi[mi.size() - 1].first;</span></span>
<span id="L18"><span class="lineNum"> 18</span> <span class="tlaGNC"> 143 : auto mi = std::vector &lt;std::pair&lt;int, float &gt;&gt;();</span></span>
<span id="L19"><span class="lineNum"> 19</span> <span class="tlaGNC"> 429 : torch::Tensor class_dataset = dataset.index({ -1, &quot;...&quot; });</span></span>
<span id="L20"><span class="lineNum"> 20</span> <span class="tlaGNC"> 979 : for (int i = 0; i &lt; static_cast&lt;int&gt;(features.size()); ++i) {</span></span>
<span id="L21"><span class="lineNum"> 21</span> <span class="tlaGNC"> 2508 : torch::Tensor feature_dataset = dataset.index({ i, &quot;...&quot; });</span></span>
<span id="L22"><span class="lineNum"> 22</span> <span class="tlaGNC"> 836 : auto mi_value = metrics.mutualInformation(class_dataset, feature_dataset, weights);</span></span>
<span id="L23"><span class="lineNum"> 23</span> <span class="tlaGNC"> 836 : mi.push_back({ i, mi_value });</span></span>
<span id="L24"><span class="lineNum"> 24</span> <span class="tlaGNC"> 836 : }</span></span>
<span id="L25"><span class="lineNum"> 25</span> <span class="tlaGNC"> 1925 : sort(mi.begin(), mi.end(), [](const auto&amp; left, const auto&amp; right) {return left.second &lt; right.second;});</span></span>
<span id="L26"><span class="lineNum"> 26</span> <span class="tlaGNC"> 143 : auto root = mi[mi.size() - 1].first;</span></span>
<span id="L27"><span class="lineNum"> 27</span> : // 2. Compute mutual information between each feature and the class</span>
<span id="L28"><span class="lineNum"> 28</span> <span class="tlaGNC"> 13 : auto weights_matrix = metrics.conditionalEdge(weights);</span></span>
<span id="L28"><span class="lineNum"> 28</span> <span class="tlaGNC"> 143 : auto weights_matrix = metrics.conditionalEdge(weights);</span></span>
<span id="L29"><span class="lineNum"> 29</span> : // 3. Compute the maximum spanning tree</span>
<span id="L30"><span class="lineNum"> 30</span> <span class="tlaGNC"> 13 : auto mst = metrics.maximumSpanningTree(features, weights_matrix, root);</span></span>
<span id="L30"><span class="lineNum"> 30</span> <span class="tlaGNC"> 143 : auto mst = metrics.maximumSpanningTree(features, weights_matrix, root);</span></span>
<span id="L31"><span class="lineNum"> 31</span> : // 4. Add edges from the maximum spanning tree to the model</span>
<span id="L32"><span class="lineNum"> 32</span> <span class="tlaGNC"> 76 : for (auto i = 0; i &lt; mst.size(); ++i) {</span></span>
<span id="L33"><span class="lineNum"> 33</span> <span class="tlaGNC"> 63 : auto [from, to] = mst[i];</span></span>
<span id="L34"><span class="lineNum"> 34</span> <span class="tlaGNC"> 63 : model.addEdge(features[from], features[to]);</span></span>
<span id="L32"><span class="lineNum"> 32</span> <span class="tlaGNC"> 836 : for (auto i = 0; i &lt; mst.size(); ++i) {</span></span>
<span id="L33"><span class="lineNum"> 33</span> <span class="tlaGNC"> 693 : auto [from, to] = mst[i];</span></span>
<span id="L34"><span class="lineNum"> 34</span> <span class="tlaGNC"> 693 : model.addEdge(features[from], features[to]);</span></span>
<span id="L35"><span class="lineNum"> 35</span> : }</span>
<span id="L36"><span class="lineNum"> 36</span> : // 5. Add edges from the class to all features</span>
<span id="L37"><span class="lineNum"> 37</span> <span class="tlaGNC"> 89 : for (auto feature : features) {</span></span>
<span id="L38"><span class="lineNum"> 38</span> <span class="tlaGNC"> 76 : model.addEdge(className, feature);</span></span>
<span id="L39"><span class="lineNum"> 39</span> <span class="tlaGNC"> 76 : }</span></span>
<span id="L40"><span class="lineNum"> 40</span> <span class="tlaGNC"> 102 : }</span></span>
<span id="L41"><span class="lineNum"> 41</span> <span class="tlaGNC"> 2 : std::vector&lt;std::string&gt; TAN::graph(const std::string&amp; title) const</span></span>
<span id="L37"><span class="lineNum"> 37</span> <span class="tlaGNC"> 979 : for (auto feature : features) {</span></span>
<span id="L38"><span class="lineNum"> 38</span> <span class="tlaGNC"> 836 : model.addEdge(className, feature);</span></span>
<span id="L39"><span class="lineNum"> 39</span> <span class="tlaGNC"> 836 : }</span></span>
<span id="L40"><span class="lineNum"> 40</span> <span class="tlaGNC"> 1122 : }</span></span>
<span id="L41"><span class="lineNum"> 41</span> <span class="tlaGNC"> 22 : std::vector&lt;std::string&gt; TAN::graph(const std::string&amp; title) const</span></span>
<span id="L42"><span class="lineNum"> 42</span> : {</span>
<span id="L43"><span class="lineNum"> 43</span> <span class="tlaGNC"> 2 : return model.graph(title);</span></span>
<span id="L43"><span class="lineNum"> 43</span> <span class="tlaGNC"> 22 : return model.graph(title);</span></span>
<span id="L44"><span class="lineNum"> 44</span> : }</span>
<span id="L45"><span class="lineNum"> 45</span> : }</span>
</pre>