P281424/AMR
1
0
Fork 0
mirror of https://github.com/msberends/AMR.git synced 2025-07-08 09:51:48 +02:00
Code Releases Activity

small as.mo fix

Browse Source
This commit is contained in:
dr. M.S. (Matthijs) Berends
2019-03-06 14:39:02 +01:00
parent 5c7a061a5d
commit 0c0c6e289b
15 changed files with 212 additions and 150 deletions
Download Patch File Download Diff File Expand all files Collapse all files

78
docs/articles/benchmarks.html
View File

@ -192,7 +192,7 @@
<h1>Benchmarks</h1>
<h4 class="author">Matthijs S. Berends</h4>
<h4 class="date">05 March 2019</h4>
<h4 class="date">06 March 2019</h4>
<div class="hidden name"><code>benchmarks.Rmd</code></div>
@ -201,12 +201,12 @@
<p>One of the most important features of this package is the complete microbial taxonomic database, supplied by the <a href="http://catalogueoflife.org">Catalogue of Life</a>. We created a function <code><a href="../reference/as.mo.html">as.mo()</a></code> that transforms any user input value to a valid microbial ID by using AI (Artificial Intelligence) combined with the taxonomic tree of Catalogue of Life.</p>
<p>One of the most important features of this package is the complete microbial taxonomic database, supplied by the <a href="http://catalogueoflife.org">Catalogue of Life</a>. We created a function <code><a href="../reference/as.mo.html">as.mo()</a></code> that transforms any user input value to a valid microbial ID by using intelligent rules combined with the taxonomic tree of Catalogue of Life.</p>
<p>Using the <code>microbenchmark</code> package, we can review the calculation performance of this function. Its function <code><a href="https://www.rdocumentation.org/packages/microbenchmark/topics/microbenchmark">microbenchmark()</a></code> runs different input expressions independently of each other and measures their time-to-result.</p>
<div class="sourceCode" id="cb1"><pre class="sourceCode r"><code class="sourceCode r"><a class="sourceLine" id="cb1-1" title="1"><span class="kw"><a href="https://www.rdocumentation.org/packages/base/topics/library">library</a></span>(microbenchmark)</a>
<a class="sourceLine" id="cb1-2" title="2"><span class="kw"><a href="https://www.rdocumentation.org/packages/base/topics/library">library</a></span>(AMR)</a></code></pre></div>
<p>In the next test, we try to ‘coerce’ different input values for <em>Staphylococcus aureus</em>. The actual result is the same every time: it returns its MO code <code>B_STPHY_AUR</code> (<em>B</em> stands for <em>Bacteria</em>, the taxonomic kingdom).</p>
<p>But the calculation time differs a lot. Here, the AI effect can be reviewed best:</p>
<p>But the calculation time differs a lot:</p>
<div class="sourceCode" id="cb2"><pre class="sourceCode r"><code class="sourceCode r"><a class="sourceLine" id="cb2-1" title="1">S.aureus &lt;-<span class="st"> </span><span class="kw"><a href="https://www.rdocumentation.org/packages/microbenchmark/topics/microbenchmark">microbenchmark</a></span>(<span class="kw"><a href="../reference/as.mo.html">as.mo</a></span>(<span class="st">"sau"</span>),</a>
<a class="sourceLine" id="cb2-2" title="2"> <span class="kw"><a href="../reference/as.mo.html">as.mo</a></span>(<span class="st">"stau"</span>),</a>
<a class="sourceLine" id="cb2-3" title="3"> <span class="kw"><a href="../reference/as.mo.html">as.mo</a></span>(<span class="st">"staaur"</span>),</a>
@ -217,14 +217,14 @@
<a class="sourceLine" id="cb2-8" title="8"> <span class="dt">times =</span> <span class="dv">10</span>)</a>
<a class="sourceLine" id="cb2-9" title="9"><span class="kw"><a href="https://www.rdocumentation.org/packages/base/topics/print">print</a></span>(S.aureus, <span class="dt">unit =</span> <span class="st">"ms"</span>, <span class="dt">signif =</span> <span class="dv">3</span>)</a>
<a class="sourceLine" id="cb2-10" title="10"><span class="co">#&gt; Unit: milliseconds</span></a>
<a class="sourceLine" id="cb2-11" title="11"><span class="co">#&gt; expr min lq mean median uq max neval</span></a>
<a class="sourceLine" id="cb2-12" title="12"><span class="co">#&gt; as.mo("sau") 16.80 17.20 25.80 18.10 32.10 52.8 10</span></a>
<a class="sourceLine" id="cb2-13" title="13"><span class="co">#&gt; as.mo("stau") 32.20 33.10 43.30 34.00 42.40 82.3 10</span></a>
<a class="sourceLine" id="cb2-14" title="14"><span class="co">#&gt; as.mo("staaur") 16.70 16.80 25.40 17.20 19.80 60.4 10</span></a>
<a class="sourceLine" id="cb2-15" title="15"><span class="co">#&gt; as.mo("STAAUR") 16.70 16.80 22.40 17.90 20.20 61.7 10</span></a>
<a class="sourceLine" id="cb2-16" title="16"><span class="co">#&gt; as.mo("S. aureus") 24.90 25.10 42.00 26.20 67.50 88.8 10</span></a>
<a class="sourceLine" id="cb2-17" title="17"><span class="co">#&gt; as.mo("S. aureus") 24.50 25.00 43.70 25.30 67.90 78.2 10</span></a>
<a class="sourceLine" id="cb2-18" title="18"><span class="co">#&gt; as.mo("Staphylococcus aureus") 7.47 7.48 9.63 7.53 7.69 28.3 10</span></a></code></pre></div>
<a class="sourceLine" id="cb2-11" title="11"><span class="co">#&gt; expr min lq mean median uq max neval</span></a>
<a class="sourceLine" id="cb2-12" title="12"><span class="co">#&gt; as.mo("sau") 16.50 16.60 17.0 16.70 17.00 19.0 10</span></a>
<a class="sourceLine" id="cb2-13" title="13"><span class="co">#&gt; as.mo("stau") 31.70 31.90 51.5 32.10 49.10 166.0 10</span></a>
<a class="sourceLine" id="cb2-14" title="14"><span class="co">#&gt; as.mo("staaur") 16.60 16.70 21.8 16.80 17.00 65.7 10</span></a>
<a class="sourceLine" id="cb2-15" title="15"><span class="co">#&gt; as.mo("STAAUR") 16.70 16.70 27.1 16.80 17.60 77.3 10</span></a>
<a class="sourceLine" id="cb2-16" title="16"><span class="co">#&gt; as.mo("S. aureus") 24.60 24.70 30.2 25.30 31.80 57.8 10</span></a>
<a class="sourceLine" id="cb2-17" title="17"><span class="co">#&gt; as.mo("S. aureus") 24.60 24.60 37.7 25.10 66.00 67.4 10</span></a>
<a class="sourceLine" id="cb2-18" title="18"><span class="co">#&gt; as.mo("Staphylococcus aureus") 7.47 7.48 11.1 7.65 8.02 41.1 10</span></a></code></pre></div>
<p>In the table above, all measurements are shown in milliseconds (thousands of seconds). A value of 5 milliseconds means it can determine 200 input values per second. It case of 100 milliseconds, this is only 10 input values per second. The second input is the only one that has to be looked up thoroughly. All the others are known codes (the first one is a WHONET code) or common laboratory codes, or common full organism names like the last one. Full organism names are always preferred.</p>
<p>To achieve this speed, the <code>as.mo</code> function also takes into account the prevalence of human pathogenic microorganisms. The downside is of course that less prevalent microorganisms will be determined less fast. See this example for the ID of <em>Thermus islandicus</em> (<code>B_THERMS_ISL</code>), a bug probably never found before in humans:</p>
<div class="sourceCode" id="cb3"><pre class="sourceCode r"><code class="sourceCode r"><a class="sourceLine" id="cb3-1" title="1">T.islandicus &lt;-<span class="st"> </span><span class="kw"><a href="https://www.rdocumentation.org/packages/microbenchmark/topics/microbenchmark">microbenchmark</a></span>(<span class="kw"><a href="../reference/as.mo.html">as.mo</a></span>(<span class="st">"theisl"</span>),</a>
@ -235,13 +235,13 @@
<a class="sourceLine" id="cb3-6" title="6"> <span class="dt">times =</span> <span class="dv">10</span>)</a>
<a class="sourceLine" id="cb3-7" title="7"><span class="kw"><a href="https://www.rdocumentation.org/packages/base/topics/print">print</a></span>(T.islandicus, <span class="dt">unit =</span> <span class="st">"ms"</span>, <span class="dt">signif =</span> <span class="dv">3</span>)</a>
<a class="sourceLine" id="cb3-8" title="8"><span class="co">#&gt; Unit: milliseconds</span></a>
<a class="sourceLine" id="cb3-9" title="9"><span class="co">#&gt; expr min lq mean median uq max neval</span></a>
<a class="sourceLine" id="cb3-10" title="10"><span class="co">#&gt; as.mo("theisl") 275.0 304.0 310 309.0 316 337 10</span></a>
<a class="sourceLine" id="cb3-11" title="11"><span class="co">#&gt; as.mo("THEISL") 263.0 304.0 311 306.0 312 391 10</span></a>
<a class="sourceLine" id="cb3-12" title="12"><span class="co">#&gt; as.mo("T. islandicus") 142.0 143.0 168 148.0 188 220 10</span></a>
<a class="sourceLine" id="cb3-13" title="13"><span class="co">#&gt; as.mo("T. islandicus") 142.0 142.0 169 143.0 185 312 10</span></a>
<a class="sourceLine" id="cb3-14" title="14"><span class="co">#&gt; as.mo("Thermus islandicus") 68.1 68.6 101 89.8 122 179 10</span></a></code></pre></div>
<p>That takes 7 times as much time on average. A value of 100 milliseconds means it can only determine ~10 different input values per second. We can conclude that looking up arbitrary codes of less prevalent microorganisms is the worst way to go, in terms of calculation performance. Full names (like <em>Thermus islandicus</em>) are almost fast - these are the most probable input from most data sets.</p>
<a class="sourceLine" id="cb3-9" title="9"><span class="co">#&gt; expr min lq mean median uq max neval</span></a>
<a class="sourceLine" id="cb3-10" title="10"><span class="co">#&gt; as.mo("theisl") 269.0 270.0 294.0 293.0 317 320 10</span></a>
<a class="sourceLine" id="cb3-11" title="11"><span class="co">#&gt; as.mo("THEISL") 272.0 313.0 327.0 316.0 321 476 10</span></a>
<a class="sourceLine" id="cb3-12" title="12"><span class="co">#&gt; as.mo("T. islandicus") 142.0 142.0 159.0 144.0 191 205 10</span></a>
<a class="sourceLine" id="cb3-13" title="13"><span class="co">#&gt; as.mo("T. islandicus") 142.0 143.0 166.0 164.0 188 196 10</span></a>
<a class="sourceLine" id="cb3-14" title="14"><span class="co">#&gt; as.mo("Thermus islandicus") 68.4 68.6 86.8 69.2 113 116 10</span></a></code></pre></div>
<p>That takes 7.4 times as much time on average. A value of 100 milliseconds means it can only determine ~10 different input values per second. We can conclude that looking up arbitrary codes of less prevalent microorganisms is the worst way to go, in terms of calculation performance. Full names (like <em>Thermus islandicus</em>) are almost fast - these are the most probable input from most data sets.</p>
<p>In the figure below, we compare <em>Escherichia coli</em> (which is very common) with <em>Prevotella brevis</em> (which is moderately common) and with <em>Thermus islandicus</em> (which is very uncommon):</p>
<div class="sourceCode" id="cb4"><pre class="sourceCode r"><code class="sourceCode r"><a class="sourceLine" id="cb4-1" title="1"><span class="kw"><a href="https://www.rdocumentation.org/packages/graphics/topics/par">par</a></span>(<span class="dt">mar =</span> <span class="kw"><a href="https://www.rdocumentation.org/packages/base/topics/c">c</a></span>(<span class="dv">5</span>, <span class="dv">16</span>, <span class="dv">4</span>, <span class="dv">2</span>)) <span class="co"># set more space for left margin text (16)</span></a>
<a class="sourceLine" id="cb4-2" title="2"></a>
@ -287,8 +287,8 @@
<a class="sourceLine" id="cb5-24" title="24"><span class="kw"><a href="https://www.rdocumentation.org/packages/base/topics/print">print</a></span>(run_it, <span class="dt">unit =</span> <span class="st">"ms"</span>, <span class="dt">signif =</span> <span class="dv">3</span>)</a>
<a class="sourceLine" id="cb5-25" title="25"><span class="co">#&gt; Unit: milliseconds</span></a>
<a class="sourceLine" id="cb5-26" title="26"><span class="co">#&gt; expr min lq mean median uq max neval</span></a>
<a class="sourceLine" id="cb5-27" title="27"><span class="co">#&gt; mo_fullname(x) 687 738 767 770 774 887 10</span></a></code></pre></div>
<p>So transforming 500,000 values (!!) of 50 unique values only takes 0.77 seconds (770 ms). You only lose time on your unique input values.</p>
<a class="sourceLine" id="cb5-27" title="27"><span class="co">#&gt; mo_fullname(x) 688 757 800 758 919 921 10</span></a></code></pre></div>
<p>So transforming 500,000 values (!!) of 50 unique values only takes 0.76 seconds (758 ms). You only lose time on your unique input values.</p>
</div>
<div id="precalculated-results" class="section level3">
<h3 class="hasAnchor">
@ -300,10 +300,10 @@
<a class="sourceLine" id="cb6-4" title="4"> <span class="dt">times =</span> <span class="dv">10</span>)</a>
<a class="sourceLine" id="cb6-5" title="5"><span class="kw"><a href="https://www.rdocumentation.org/packages/base/topics/print">print</a></span>(run_it, <span class="dt">unit =</span> <span class="st">"ms"</span>, <span class="dt">signif =</span> <span class="dv">3</span>)</a>
<a class="sourceLine" id="cb6-6" title="6"><span class="co">#&gt; Unit: milliseconds</span></a>
<a class="sourceLine" id="cb6-7" title="7"><span class="co">#&gt; expr min lq mean median uq max neval</span></a>
<a class="sourceLine" id="cb6-8" title="8"><span class="co">#&gt; A 11.100 11.300 15.600 11.400 11.600 53.4 10</span></a>
<a class="sourceLine" id="cb6-9" title="9"><span class="co">#&gt; B 22.200 22.500 22.800 22.800 23.000 23.7 10</span></a>
<a class="sourceLine" id="cb6-10" title="10"><span class="co">#&gt; C 0.322 0.326 0.518 0.563 0.578 0.8 10</span></a></code></pre></div>
<a class="sourceLine" id="cb6-7" title="7"><span class="co">#&gt; expr min lq mean median uq max neval</span></a>
<a class="sourceLine" id="cb6-8" title="8"><span class="co">#&gt; A 11.200 11.300 11.900 12.200 12.300 12.600 10</span></a>
<a class="sourceLine" id="cb6-9" title="9"><span class="co">#&gt; B 22.300 22.500 23.300 23.100 23.900 24.700 10</span></a>
<a class="sourceLine" id="cb6-10" title="10"><span class="co">#&gt; C 0.339 0.519 0.612 0.564 0.768 0.776 10</span></a></code></pre></div>
<p>So going from <code><a href="../reference/mo_property.html">mo_fullname("Staphylococcus aureus")</a></code> to <code>"Staphylococcus aureus"</code> takes 0.0006 seconds - it doesn’t even start calculating <em>if the result would be the same as the expected resulting value</em>. That goes for all helper functions:</p>
<div class="sourceCode" id="cb7"><pre class="sourceCode r"><code class="sourceCode r"><a class="sourceLine" id="cb7-1" title="1">run_it &lt;-<span class="st"> </span><span class="kw"><a href="https://www.rdocumentation.org/packages/microbenchmark/topics/microbenchmark">microbenchmark</a></span>(<span class="dt">A =</span> <span class="kw"><a href="../reference/mo_property.html">mo_species</a></span>(<span class="st">"aureus"</span>),</a>
<a class="sourceLine" id="cb7-2" title="2"> <span class="dt">B =</span> <span class="kw"><a href="../reference/mo_property.html">mo_genus</a></span>(<span class="st">"Staphylococcus"</span>),</a>
@ -317,14 +317,14 @@
<a class="sourceLine" id="cb7-10" title="10"><span class="kw"><a href="https://www.rdocumentation.org/packages/base/topics/print">print</a></span>(run_it, <span class="dt">unit =</span> <span class="st">"ms"</span>, <span class="dt">signif =</span> <span class="dv">3</span>)</a>
<a class="sourceLine" id="cb7-11" title="11"><span class="co">#&gt; Unit: milliseconds</span></a>
<a class="sourceLine" id="cb7-12" title="12"><span class="co">#&gt; expr min lq mean median uq max neval</span></a>
<a class="sourceLine" id="cb7-13" title="13"><span class="co">#&gt; A 0.330 0.345 0.421 0.441 0.461 0.525 10</span></a>
<a class="sourceLine" id="cb7-14" title="14"><span class="co">#&gt; B 0.355 0.378 0.453 0.435 0.464 0.754 10</span></a>
<a class="sourceLine" id="cb7-15" title="15"><span class="co">#&gt; C 0.315 0.374 0.506 0.516 0.633 0.700 10</span></a>
<a class="sourceLine" id="cb7-16" title="16"><span class="co">#&gt; D 0.273 0.329 0.353 0.365 0.385 0.391 10</span></a>
<a class="sourceLine" id="cb7-17" title="17"><span class="co">#&gt; E 0.279 0.322 0.330 0.335 0.343 0.366 10</span></a>
<a class="sourceLine" id="cb7-18" title="18"><span class="co">#&gt; F 0.247 0.263 0.315 0.319 0.355 0.429 10</span></a>
<a class="sourceLine" id="cb7-19" title="19"><span class="co">#&gt; G 0.276 0.320 0.342 0.328 0.371 0.447 10</span></a>
<a class="sourceLine" id="cb7-20" title="20"><span class="co">#&gt; H 0.251 0.258 0.315 0.329 0.360 0.368 10</span></a></code></pre></div>
<a class="sourceLine" id="cb7-13" title="13"><span class="co">#&gt; A 0.321 0.434 0.481 0.501 0.519 0.665 10</span></a>
<a class="sourceLine" id="cb7-14" title="14"><span class="co">#&gt; B 0.352 0.414 0.482 0.475 0.513 0.715 10</span></a>
<a class="sourceLine" id="cb7-15" title="15"><span class="co">#&gt; C 0.394 0.648 0.670 0.679 0.770 0.839 10</span></a>
<a class="sourceLine" id="cb7-16" title="16"><span class="co">#&gt; D 0.354 0.371 0.393 0.395 0.410 0.428 10</span></a>
<a class="sourceLine" id="cb7-17" title="17"><span class="co">#&gt; E 0.286 0.353 0.370 0.369 0.399 0.443 10</span></a>
<a class="sourceLine" id="cb7-18" title="18"><span class="co">#&gt; F 0.317 0.373 0.377 0.380 0.392 0.437 10</span></a>
<a class="sourceLine" id="cb7-19" title="19"><span class="co">#&gt; G 0.272 0.307 0.352 0.348 0.387 0.431 10</span></a>
<a class="sourceLine" id="cb7-20" title="20"><span class="co">#&gt; H 0.293 0.338 0.366 0.361 0.412 0.441 10</span></a></code></pre></div>
<p>Of course, when running <code><a href="../reference/mo_property.html">mo_phylum("Firmicutes")</a></code> the function has zero knowledge about the actual microorganism, namely <em>S. aureus</em>. But since the result would be <code>"Firmicutes"</code> too, there is no point in calculating the result. And because this package ‘knows’ all phyla of all known bacteria (according to the Catalogue of Life), it can just return the initial value immediately.</p>
</div>
<div id="results-in-other-languages" class="section level3">
@ -351,13 +351,13 @@
<a class="sourceLine" id="cb8-18" title="18"><span class="kw"><a href="https://www.rdocumentation.org/packages/base/topics/print">print</a></span>(run_it, <span class="dt">unit =</span> <span class="st">"ms"</span>, <span class="dt">signif =</span> <span class="dv">4</span>)</a>
<a class="sourceLine" id="cb8-19" title="19"><span class="co">#&gt; Unit: milliseconds</span></a>
<a class="sourceLine" id="cb8-20" title="20"><span class="co">#&gt; expr min lq mean median uq max neval</span></a>
<a class="sourceLine" id="cb8-21" title="21"><span class="co">#&gt; en 14.98 15.35 19.58 15.51 15.58 56.78 10</span></a>
<a class="sourceLine" id="cb8-22" title="22"><span class="co">#&gt; de 27.26 27.69 28.05 27.85 28.01 29.44 10</span></a>
<a class="sourceLine" id="cb8-23" title="23"><span class="co">#&gt; nl 27.07 27.73 31.89 27.87 28.00 69.50 10</span></a>
<a class="sourceLine" id="cb8-24" title="24"><span class="co">#&gt; es 26.98 27.43 31.98 27.84 28.09 69.30 10</span></a>
<a class="sourceLine" id="cb8-25" title="25"><span class="co">#&gt; it 27.26 27.57 28.38 27.82 28.92 32.14 10</span></a>
<a class="sourceLine" id="cb8-26" title="26"><span class="co">#&gt; fr 27.21 27.72 36.16 27.83 28.19 70.08 10</span></a>
<a class="sourceLine" id="cb8-27" title="27"><span class="co">#&gt; pt 27.21 27.78 36.17 27.96 28.54 69.43 10</span></a></code></pre></div>
<a class="sourceLine" id="cb8-21" title="21"><span class="co">#&gt; en 14.97 15.49 15.46 15.56 15.60 15.65 10</span></a>
<a class="sourceLine" id="cb8-22" title="22"><span class="co">#&gt; de 27.29 27.76 44.46 27.98 69.27 71.00 10</span></a>
<a class="sourceLine" id="cb8-23" title="23"><span class="co">#&gt; nl 27.22 27.97 28.53 28.03 29.40 31.36 10</span></a>
<a class="sourceLine" id="cb8-24" title="24"><span class="co">#&gt; es 27.45 27.96 32.24 28.00 28.08 69.99 10</span></a>
<a class="sourceLine" id="cb8-25" title="25"><span class="co">#&gt; it 27.21 27.32 28.01 27.89 28.05 29.41 10</span></a>
<a class="sourceLine" id="cb8-26" title="26"><span class="co">#&gt; fr 27.26 27.94 28.16 28.12 28.25 29.47 10</span></a>
<a class="sourceLine" id="cb8-27" title="27"><span class="co">#&gt; pt 27.08 27.45 36.14 27.92 28.07 70.34 10</span></a></code></pre></div>
<p>Currently supported are German, Dutch, Spanish, Italian, French and Portuguese.</p>
</div>
</div>