rmontanana/BayesNet
rmontanana/BayesNet
1
0
Fork 0
Code Issues 2 Pull Requests Packages Projects Releases 3 Wiki Activity
AnDE
BayesNet/html/bayesnet/feature_selection/FeatureSelect.cc.gcov.html
Ricardo Montañana Gómez ced29a2c2e
Refactor coverage report generation 
Add some tests to reach 99%
2024-05-06 17:56:00 +02:00

169 lines
14 KiB
HTML
Raw Permalink Blame History

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html lang="en">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<title>LCOV - BayesNet Coverage Report - bayesnet/feature_selection/FeatureSelect.cc</title>
<link rel="stylesheet" type="text/css" href="../../gcov.css">
</head>
<body>
<table width="100%" border=0 cellspacing=0 cellpadding=0>
<tr><td class="title">LCOV - code coverage report</td></tr>
<tr><td class="ruler"><img src="../../glass.png" width=3 height=3 alt=""></td></tr>
<tr>
<td width="100%">
<table cellpadding=1 border=0 width="100%">
<tr>
<td width="10%" class="headerItem">Current view:</td>
<td width="10%" class="headerValue"><a href="../../index.html" target="_parent">top level</a> - <a href="index.html" target="_parent">bayesnet/feature_selection</a> - FeatureSelect.cc<span style="font-size: 80%;"> (source / <a href="FeatureSelect.cc.func-c.html">functions</a>)</span></td>
<td width="5%"></td>
<td width="5%"></td>
<td width="5%" class="headerCovTableHead">Coverage</td>
<td width="5%" class="headerCovTableHead" title="Covered + Uncovered code">Total</td>
<td width="5%" class="headerCovTableHead" title="Exercised code only">Hit</td>
</tr>
<tr>
<td class="headerItem">Test:</td>
<td class="headerValue">BayesNet Coverage Report</td>
<td></td>
<td class="headerItem">Lines:</td>
<td class="headerCovTableEntryHi">100.0&nbsp;%</td>
<td class="headerCovTableEntry">44</td>
<td class="headerCovTableEntry">44</td>
</tr>
<tr>
<td class="headerItem">Test Date:</td>
<td class="headerValue">2024-05-06 17:54:04</td>
<td></td>
<td class="headerItem">Functions:</td>
<td class="headerCovTableEntryHi">100.0&nbsp;%</td>
<td class="headerCovTableEntry">8</td>
<td class="headerCovTableEntry">8</td>
</tr>
<tr>
<td class="headerItem">Legend:</td>
<td class="headerValueLeg"> Lines:
<span class="coverLegendCov">hit</span>
<span class="coverLegendNoCov">not hit</span>
</td>
<td></td>
</tr>
<tr><td><img src="../../glass.png" width=3 height=3 alt=""></td></tr>
</table>
</td>
</tr>
<tr><td class="ruler"><img src="../../glass.png" width=3 height=3 alt=""></td></tr>
</table>
<table cellpadding=0 cellspacing=0 border=0>
<tr>
<td><br></td>
</tr>
<tr>
<td>
<pre class="sourceHeading"> Line data Source code</pre>
<pre class="source">
<span id="L1"><span class="lineNum"> 1</span> : // ***************************************************************</span>
<span id="L2"><span class="lineNum"> 2</span> : // SPDX-FileCopyrightText: Copyright 2024 Ricardo Montañana Gómez</span>
<span id="L3"><span class="lineNum"> 3</span> : // SPDX-FileType: SOURCE</span>
<span id="L4"><span class="lineNum"> 4</span> : // SPDX-License-Identifier: MIT</span>
<span id="L5"><span class="lineNum"> 5</span> : // ***************************************************************</span>
<span id="L6"><span class="lineNum"> 6</span> : </span>
<span id="L7"><span class="lineNum"> 7</span> : #include &lt;limits&gt;</span>
<span id="L8"><span class="lineNum"> 8</span> : #include &quot;bayesnet/utils/bayesnetUtils.h&quot;</span>
<span id="L9"><span class="lineNum"> 9</span> : #include &quot;FeatureSelect.h&quot;</span>
<span id="L10"><span class="lineNum"> 10</span> : namespace bayesnet {</span>
<span id="L11"><span class="lineNum"> 11</span> <span class="tlaGNC tlaBgGNC"> 221 : FeatureSelect::FeatureSelect(const torch::Tensor&amp; samples, const std::vector&lt;std::string&gt;&amp; features, const std::string&amp; className, const int maxFeatures, const int classNumStates, const torch::Tensor&amp; weights) :</span></span>
<span id="L12"><span class="lineNum"> 12</span> <span class="tlaGNC"> 221 : Metrics(samples, features, className, classNumStates), maxFeatures(maxFeatures == 0 ? samples.size(0) - 1 : maxFeatures), weights(weights)</span></span>
<span id="L13"><span class="lineNum"> 13</span> : </span>
<span id="L14"><span class="lineNum"> 14</span> : {</span>
<span id="L15"><span class="lineNum"> 15</span> <span class="tlaGNC"> 221 : }</span></span>
<span id="L16"><span class="lineNum"> 16</span> <span class="tlaGNC"> 151 : void FeatureSelect::initialize()</span></span>
<span id="L17"><span class="lineNum"> 17</span> : {</span>
<span id="L18"><span class="lineNum"> 18</span> <span class="tlaGNC"> 151 : selectedFeatures.clear();</span></span>
<span id="L19"><span class="lineNum"> 19</span> <span class="tlaGNC"> 151 : selectedScores.clear();</span></span>
<span id="L20"><span class="lineNum"> 20</span> <span class="tlaGNC"> 151 : }</span></span>
<span id="L21"><span class="lineNum"> 21</span> <span class="tlaGNC"> 3728 : double FeatureSelect::symmetricalUncertainty(int a, int b)</span></span>
<span id="L22"><span class="lineNum"> 22</span> : {</span>
<span id="L23"><span class="lineNum"> 23</span> : /*</span>
<span id="L24"><span class="lineNum"> 24</span> : Compute symmetrical uncertainty. Normalize* information gain (mutual</span>
<span id="L25"><span class="lineNum"> 25</span> : information) with the entropies of the features in order to compensate</span>
<span id="L26"><span class="lineNum"> 26</span> : the bias due to high cardinality features. *Range [0, 1]</span>
<span id="L27"><span class="lineNum"> 27</span> : (https://www.sciencedirect.com/science/article/pii/S0020025519303603)</span>
<span id="L28"><span class="lineNum"> 28</span> : */</span>
<span id="L29"><span class="lineNum"> 29</span> <span class="tlaGNC"> 11184 : auto x = samples.index({ a, &quot;...&quot; });</span></span>
<span id="L30"><span class="lineNum"> 30</span> <span class="tlaGNC"> 11184 : auto y = samples.index({ b, &quot;...&quot; });</span></span>
<span id="L31"><span class="lineNum"> 31</span> <span class="tlaGNC"> 3728 : auto mu = mutualInformation(x, y, weights);</span></span>
<span id="L32"><span class="lineNum"> 32</span> <span class="tlaGNC"> 3728 : auto hx = entropy(x, weights);</span></span>
<span id="L33"><span class="lineNum"> 33</span> <span class="tlaGNC"> 3728 : auto hy = entropy(y, weights);</span></span>
<span id="L34"><span class="lineNum"> 34</span> <span class="tlaGNC"> 3728 : return 2.0 * mu / (hx + hy);</span></span>
<span id="L35"><span class="lineNum"> 35</span> <span class="tlaGNC"> 11184 : }</span></span>
<span id="L36"><span class="lineNum"> 36</span> <span class="tlaGNC"> 151 : void FeatureSelect::computeSuLabels()</span></span>
<span id="L37"><span class="lineNum"> 37</span> : {</span>
<span id="L38"><span class="lineNum"> 38</span> : // Compute Simmetrical Uncertainty between features and labels</span>
<span id="L39"><span class="lineNum"> 39</span> : // https://en.wikipedia.org/wiki/Symmetric_uncertainty</span>
<span id="L40"><span class="lineNum"> 40</span> <span class="tlaGNC"> 1258 : for (int i = 0; i &lt; features.size(); ++i) {</span></span>
<span id="L41"><span class="lineNum"> 41</span> <span class="tlaGNC"> 1107 : suLabels.push_back(symmetricalUncertainty(i, -1));</span></span>
<span id="L42"><span class="lineNum"> 42</span> : }</span>
<span id="L43"><span class="lineNum"> 43</span> <span class="tlaGNC"> 151 : }</span></span>
<span id="L44"><span class="lineNum"> 44</span> <span class="tlaGNC"> 7921 : double FeatureSelect::computeSuFeatures(const int firstFeature, const int secondFeature)</span></span>
<span id="L45"><span class="lineNum"> 45</span> : {</span>
<span id="L46"><span class="lineNum"> 46</span> : // Compute Simmetrical Uncertainty between features</span>
<span id="L47"><span class="lineNum"> 47</span> : // https://en.wikipedia.org/wiki/Symmetric_uncertainty</span>
<span id="L48"><span class="lineNum"> 48</span> : try {</span>
<span id="L49"><span class="lineNum"> 49</span> <span class="tlaGNC"> 7921 : return suFeatures.at({ firstFeature, secondFeature });</span></span>
<span id="L50"><span class="lineNum"> 50</span> : }</span>
<span id="L51"><span class="lineNum"> 51</span> <span class="tlaGNC"> 2621 : catch (const std::out_of_range&amp; e) {</span></span>
<span id="L52"><span class="lineNum"> 52</span> <span class="tlaGNC"> 2621 : double result = symmetricalUncertainty(firstFeature, secondFeature);</span></span>
<span id="L53"><span class="lineNum"> 53</span> <span class="tlaGNC"> 2621 : suFeatures[{firstFeature, secondFeature}] = result;</span></span>
<span id="L54"><span class="lineNum"> 54</span> <span class="tlaGNC"> 2621 : return result;</span></span>
<span id="L55"><span class="lineNum"> 55</span> <span class="tlaGNC"> 2621 : }</span></span>
<span id="L56"><span class="lineNum"> 56</span> : }</span>
<span id="L57"><span class="lineNum"> 57</span> <span class="tlaGNC"> 1239 : double FeatureSelect::computeMeritCFS()</span></span>
<span id="L58"><span class="lineNum"> 58</span> : {</span>
<span id="L59"><span class="lineNum"> 59</span> <span class="tlaGNC"> 1239 : double rcf = 0;</span></span>
<span id="L60"><span class="lineNum"> 60</span> <span class="tlaGNC"> 5693 : for (auto feature : selectedFeatures) {</span></span>
<span id="L61"><span class="lineNum"> 61</span> <span class="tlaGNC"> 4454 : rcf += suLabels[feature];</span></span>
<span id="L62"><span class="lineNum"> 62</span> : }</span>
<span id="L63"><span class="lineNum"> 63</span> <span class="tlaGNC"> 1239 : double rff = 0;</span></span>
<span id="L64"><span class="lineNum"> 64</span> <span class="tlaGNC"> 1239 : int n = selectedFeatures.size();</span></span>
<span id="L65"><span class="lineNum"> 65</span> <span class="tlaGNC"> 8154 : for (const auto&amp; item : doCombinations(selectedFeatures)) {</span></span>
<span id="L66"><span class="lineNum"> 66</span> <span class="tlaGNC"> 6915 : rff += computeSuFeatures(item.first, item.second);</span></span>
<span id="L67"><span class="lineNum"> 67</span> <span class="tlaGNC"> 1239 : }</span></span>
<span id="L68"><span class="lineNum"> 68</span> <span class="tlaGNC"> 1239 : return rcf / sqrt(n + (n * n - n) * rff);</span></span>
<span id="L69"><span class="lineNum"> 69</span> : }</span>
<span id="L70"><span class="lineNum"> 70</span> <span class="tlaGNC"> 171 : std::vector&lt;int&gt; FeatureSelect::getFeatures() const</span></span>
<span id="L71"><span class="lineNum"> 71</span> : {</span>
<span id="L72"><span class="lineNum"> 72</span> <span class="tlaGNC"> 171 : if (!fitted) {</span></span>
<span id="L73"><span class="lineNum"> 73</span> <span class="tlaGNC"> 20 : throw std::runtime_error(&quot;FeatureSelect not fitted&quot;);</span></span>
<span id="L74"><span class="lineNum"> 74</span> : }</span>
<span id="L75"><span class="lineNum"> 75</span> <span class="tlaGNC"> 151 : return selectedFeatures;</span></span>
<span id="L76"><span class="lineNum"> 76</span> : }</span>
<span id="L77"><span class="lineNum"> 77</span> <span class="tlaGNC"> 156 : std::vector&lt;double&gt; FeatureSelect::getScores() const</span></span>
<span id="L78"><span class="lineNum"> 78</span> : {</span>
<span id="L79"><span class="lineNum"> 79</span> <span class="tlaGNC"> 156 : if (!fitted) {</span></span>
<span id="L80"><span class="lineNum"> 80</span> <span class="tlaGNC"> 20 : throw std::runtime_error(&quot;FeatureSelect not fitted&quot;);</span></span>
<span id="L81"><span class="lineNum"> 81</span> : }</span>
<span id="L82"><span class="lineNum"> 82</span> <span class="tlaGNC"> 136 : return selectedScores;</span></span>
<span id="L83"><span class="lineNum"> 83</span> : }</span>
<span id="L84"><span class="lineNum"> 84</span> : }</span>
</pre>
</td>
</tr>
</table>
<br>
<table width="100%" border=0 cellspacing=0 cellpadding=0>
<tr><td class="ruler"><img src="../../glass.png" width=3 height=3 alt=""></td></tr>
<tr><td class="versionInfo">Generated by: <a href="https://github.com//linux-test-project/lcov" target="_parent">LCOV version 2.0-1</a></td></tr>
</table>
<br>
</body>
</html>
Reference in New Issue View Git Blame Copy Permalink