/usr/share/doc/python-pebl/html/tutorial.html is in python-pebl-doc 1.0.2-2.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 | <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<title>Tutorial — Pebl v1.0.1 documentation</title>
<link rel="stylesheet" href="_static/default.css" type="text/css" />
<link rel="stylesheet" href="_static/pygments.css" type="text/css" />
<script type="text/javascript">
var DOCUMENTATION_OPTIONS = {
URL_ROOT: '',
VERSION: '1.0.1',
COLLAPSE_INDEX: false,
FILE_SUFFIX: '.html',
HAS_SOURCE: true
};
</script>
<script type="text/javascript" src="_static/jquery.js"></script>
<script type="text/javascript" src="_static/underscore.js"></script>
<script type="text/javascript" src="_static/doctools.js"></script>
<link rel="top" title="Pebl v1.0.1 documentation" href="index.html" />
<link rel="next" title="Developer’s Guide" href="devguide.html" />
<link rel="prev" title="Installation" href="install.html" />
</head>
<body>
<div class="related">
<h3>Navigation</h3>
<ul>
<li class="right" style="margin-right: 10px">
<a href="genindex.html" title="General Index"
accesskey="I">index</a></li>
<li class="right" >
<a href="py-modindex.html" title="Python Module Index"
>modules</a> |</li>
<li class="right" >
<a href="devguide.html" title="Developer’s Guide"
accesskey="N">next</a> |</li>
<li class="right" >
<a href="install.html" title="Installation"
accesskey="P">previous</a> |</li>
<li><a href="index.html">Pebl v1.0.1 documentation</a> »</li>
</ul>
</div>
<div class="document">
<div class="documentwrapper">
<div class="bodywrapper">
<div class="body">
<div class="section" id="tutorial">
<span id="id1"></span><h1>Tutorial<a class="headerlink" href="#tutorial" title="Permalink to this headline">¶</a></h1>
<p>This tutorial is organized as a series of examples that highlight some of
Pebl’s various features. It is assumed that the reader is familiar with
Bayesian networks and the python language and has read the <a class="reference internal" href="intro.html#intro"><em>Pebl Introduction</em></a>.</p>
<p>Pebl includes a python library and a command line application that can be used
with a configuration file. The pebl application is limited compared to the
library but requires no programming. Each example in this tutorial will
include a python script and a pebl configuration file that runs the same
analysis (when possible).</p>
<div class="admonition note">
<p class="first admonition-title">Note</p>
<p class="last">This tutorial uses Pebl 1.0</p>
</div>
<div class="section" id="introducing-the-problem">
<h2>Introducing the Problem<a class="headerlink" href="#introducing-the-problem" title="Permalink to this headline">¶</a></h2>
<p>Bayesian networks have been used to model complex phenomena with non-linear,
multimodal relationships between high-dimensional variables. When used to model
gene regulatory networks, nodes usually represent the expression profile of
genes while edges represent dependencies between them. The learned networks can
be interpreted as causal models explaining the data and can hint at the
underlying biological mechanisms and functions.</p>
<p>For this tutorial, we use the Cell Cycle data from Spellman, et. al [1] as an
example dataset. This dataset contains 76 gene expression measurements of 6177
S. cerevisiae ORFs. The experiments include six time series using different
cell cycle synchronization methods. In this example, we ignore the temporal
aspect of the dataset and treat each measurement as an independent sample from
the underlying biological phenomena. Spellman et al. identified 800 genes as
cell cycle dependent and we use a small 12 gene subset of that (as show in Fig
8.11 of [2]).</p>
<p><a class="reference external" href="_static/tutorial/pebl-tutorial-data1.txt">pebl-tutorial-data1.txt</a> contains
the gene expression measurements for our 12 genes. Each row contains one
sample (the gene expression data from one microarray assay) and each column
contains the gene expression profile for a given gene over the 76 measurements.
The first row containing the variable names (the gene associated with the
measured ORF) is required for all pebl data files.</p>
</div>
<div class="section" id="first-example">
<h2>First Example<a class="headerlink" href="#first-example" title="Permalink to this headline">¶</a></h2>
<p>This is the simplest analysis in Pebl:</p>
<div class="highlight-python"><table class="highlighttable"><tr><td class="linenos"><div class="linenodiv"><pre>1
2
3
4
5
6
7</pre></div></td><td class="code"><div class="highlight"><pre><span class="gp">>>> </span><span class="kn">from</span> <span class="nn">pebl</span> <span class="kn">import</span> <span class="n">data</span>
<span class="gp">>>> </span><span class="kn">from</span> <span class="nn">pebl.learner</span> <span class="kn">import</span> <span class="n">greedy</span>
<span class="gp">>>> </span><span class="n">dataset</span> <span class="o">=</span> <span class="n">data</span><span class="o">.</span><span class="n">fromfile</span><span class="p">(</span><span class="s">"pebl-tutorial-data1.txt"</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">dataset</span><span class="o">.</span><span class="n">discretize</span><span class="p">()</span>
<span class="gp">>>> </span><span class="n">learner</span> <span class="o">=</span> <span class="n">greedy</span><span class="o">.</span><span class="n">GreedyLearner</span><span class="p">(</span><span class="n">dataset</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">ex1result</span> <span class="o">=</span> <span class="n">learner</span><span class="o">.</span><span class="n">run</span><span class="p">()</span>
<span class="gp">>>> </span><span class="n">ex1result</span><span class="o">.</span><span class="n">tohtml</span><span class="p">(</span><span class="s">"example1-result"</span><span class="p">)</span>
</pre></div>
</td></tr></table></div>
<p>The code above does the following:</p>
<blockquote>
<div><ul class="simple">
<li>Imports the required modules (lines 1-2)</li>
<li>Loads the dataset (line 3)</li>
<li>Discretizes the continuous values in the dataset (line 4)</li>
<li>Creates and runs a greedy learner with default parameters (lines 5-6)</li>
<li>Creates a html report of the results (line 7)</li>
</ul>
</div></blockquote>
<p>The same analysis can be run with the following configuration file:</p>
<div class="highlight-python"><table class="highlighttable"><tr><td class="linenos"><div class="linenodiv"><pre> 1
2
3
4
5
6
7
8
9
10</pre></div></td><td class="code"><div class="highlight"><pre><span class="p">[</span><span class="n">data</span><span class="p">]</span>
<span class="n">filename</span> <span class="o">=</span> <span class="n">pebl</span><span class="o">-</span><span class="n">tutorial</span><span class="o">-</span><span class="n">data1</span><span class="o">.</span><span class="n">txt</span>
<span class="n">discretize</span> <span class="o">=</span> <span class="mi">3</span>
<span class="p">[</span><span class="n">learner</span><span class="p">]</span>
<span class="nb">type</span> <span class="o">=</span> <span class="n">greedy</span><span class="o">.</span><span class="n">GreedyLearner</span>
<span class="p">[</span><span class="n">result</span><span class="p">]</span>
<span class="n">format</span> <span class="o">=</span> <span class="n">html</span>
<span class="n">outdir</span> <span class="o">=</span> <span class="n">example1</span><span class="o">-</span><span class="n">result</span>
</pre></div>
</td></tr></table></div>
<p>To use the configuration file above, save it as a text file (config1.txt) and
run the pebl application:</p>
<div class="highlight-python"><pre>$/usr/local/bin/pebl run config1.txt</pre>
</div>
<div class="admonition note">
<p class="first admonition-title">Note</p>
<p class="last">The location of the pebl application may be different based on your installation method</p>
</div>
<p>The result of this example is available <a class="reference external" href="_static/tutorial/example1-result/index.html">here</a>. Keep in mind that structure
learning of Bayesian networks uses stochastic methods and the results from
different runs will not be identical. Also, the results from this short run are
not realistic or spectacular; they do, however, serve as a good demostration of
the features of the html report.</p>
<p>The result is organized intro three tabs. The first tab shows the log scores
for the top networks and some overall statistics. The second tab shows
the top 10 networks found and the last tab shows consensus networks at
different confidence thresholds. The consensus networks are built using
estimated marginal posterior probabililty of each edge.</p>
</div>
<div class="section" id="pebl-s-data-file-format">
<h2>Pebl’s Data File Format<a class="headerlink" href="#pebl-s-data-file-format" title="Permalink to this headline">¶</a></h2>
<p>Pebl uses tab-delimited text files for its data. Each column represents a
variable and each row represents a sample or observation. The data file can
contain any number of comment rows that begin with a “#”. The first
non-comment row is expected to be a header row that specifies the name and type
of each variable. Pebl supports continuous, discrete and class variables. The
three variable types have different header formats as shown below:</p>
<table border="1" class="docutils">
<colgroup>
<col width="14%" />
<col width="51%" />
<col width="36%" />
</colgroup>
<thead valign="bottom">
<tr><th class="head">Type</th>
<th class="head">Header Format</th>
<th class="head">Examples</th>
</tr>
</thead>
<tbody valign="top">
<tr><td rowspan="2">continuous</td>
<td rowspan="2"><variable-name>[,continuous]</td>
<td rowspan="2">CLN2
CLN1,continuous</td>
</tr>
<tr></tr>
<tr><td>discrete</td>
<td><variable-name>,discrete(<variable-arity>)</td>
<td>CLN2,discrete(2)
CLN1,discrete(3)</td>
</tr>
<tr><td>class</td>
<td><variable-name>,class(<class-labels>)</td>
<td>CLN2,class(low,high)
phenotype,class(cancer,normal)</td>
</tr>
</tbody>
</table>
<div class="admonition note">
<p class="first admonition-title">Note</p>
<p class="last">Although Pebl accepts continuous values in the data file, they must be discretized before use.</p>
</div>
<p>Each measured value can be annotated with two indicators. Append or prepend a
”!” to the value to indicate that the variable was intervened upon for that
sample or observation. The intervention can be a gene knockdown, RNA silencing
or any perturbation that directly affect the value for that variable. Missing
values are indicated by using “X”. This can be the result of a scratch on a
micrarray slide or, if all the rows for a variable include “X”, a variable that
wasn’t measured.</p>
<p>Each sample (row) can have a name which should be in the first column. This is
not used in learning a Bayesian network, but can be used to create subsets of
the data based on the sample names.</p>
<p><a class="reference external" href="_static/tutorial/pebl-tutorial-data2.txt">pebl-tutorial-data2.txt</a> is the
discretized version of our data file with sample names and was created with the
following script:</p>
<div class="highlight-python"><table class="highlighttable"><tr><td class="linenos"><div class="linenodiv"><pre>1
2
3
4
5
6</pre></div></td><td class="code"><div class="highlight"><pre><span class="gp">>>> </span><span class="kn">from</span> <span class="nn">pebl</span> <span class="kn">import</span> <span class="n">data</span>
<span class="gp">>>> </span><span class="n">dataset</span> <span class="o">=</span> <span class="n">data</span><span class="o">.</span><span class="n">fromfile</span><span class="p">(</span><span class="s">"pebl-tutorial-data1.txt"</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">dataset</span><span class="o">.</span><span class="n">discretize</span><span class="p">(</span><span class="n">numbins</span><span class="o">=</span><span class="mi">3</span><span class="p">)</span>
<span class="gp">>>> </span><span class="k">for</span> <span class="n">i</span><span class="p">,</span><span class="n">s</span> <span class="ow">in</span> <span class="nb">enumerate</span><span class="p">(</span><span class="n">dataset</span><span class="o">.</span><span class="n">samples</span><span class="p">):</span>
<span class="gp">>>> </span> <span class="n">s</span><span class="o">.</span><span class="n">name</span> <span class="o">=</span> <span class="s">"sample-</span><span class="si">%d</span><span class="s">"</span> <span class="o">%</span> <span class="n">i</span>
<span class="gp">>>> </span><span class="n">dataset</span><span class="o">.</span><span class="n">tofile</span><span class="p">(</span><span class="s">"pebl-tutorial-data2.txt"</span><span class="p">)</span>
</pre></div>
</td></tr></table></div>
</div>
<div class="section" id="second-example">
<h2>Second Example<a class="headerlink" href="#second-example" title="Permalink to this headline">¶</a></h2>
<p>In the first example, we used the default parameters for the greedy learner
(1000 iterations) which is inadequate for a dataset of this size. In this
example, we use custom stopping criteria:</p>
<div class="highlight-python"><table class="highlighttable"><tr><td class="linenos"><div class="linenodiv"><pre>1
2
3
4
5
6
7
8
9</pre></div></td><td class="code"><div class="highlight"><pre><span class="gp">>>> </span><span class="kn">from</span> <span class="nn">pebl</span> <span class="kn">import</span> <span class="n">data</span><span class="p">,</span> <span class="n">result</span>
<span class="gp">>>> </span><span class="kn">from</span> <span class="nn">pebl.learner</span> <span class="kn">import</span> <span class="n">greedy</span>
<span class="gp">>>> </span><span class="n">dataset</span> <span class="o">=</span> <span class="n">data</span><span class="o">.</span><span class="n">fromfile</span><span class="p">(</span><span class="s">"pebl-tutorial-data2.txt"</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">learner1</span> <span class="o">=</span> <span class="n">greedy</span><span class="o">.</span><span class="n">GreedyLearner</span><span class="p">(</span><span class="n">dataset</span><span class="p">,</span> <span class="n">max_iterations</span><span class="o">=</span><span class="mi">1000000</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">learner2</span> <span class="o">=</span> <span class="n">greedy</span><span class="o">.</span><span class="n">GreedyLearner</span><span class="p">(</span><span class="n">dataset</span><span class="p">,</span> <span class="n">max_time</span><span class="o">=</span><span class="mi">120</span><span class="p">)</span> <span class="c"># in seconds</span>
<span class="gp">>>> </span><span class="n">result1</span> <span class="o">=</span> <span class="n">learner1</span><span class="o">.</span><span class="n">run</span><span class="p">()</span>
<span class="gp">>>> </span><span class="n">result2</span> <span class="o">=</span> <span class="n">learner2</span><span class="o">.</span><span class="n">run</span><span class="p">()</span>
<span class="gp">>>> </span><span class="n">merged_result</span> <span class="o">=</span> <span class="n">result</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">result1</span><span class="p">,</span> <span class="n">result2</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">merged_result</span><span class="o">.</span><span class="n">tohtml</span><span class="p">(</span><span class="s">"example2-result"</span><span class="p">)</span>
</pre></div>
</td></tr></table></div>
<p>The code above does the following:</p>
<blockquote>
<div><ul class="simple">
<li>Imports the required modules (lines 1-2)</li>
<li>Loads the discretized dataset (line 3)</li>
<li>Creates and runs two greedy learners with specified stopping criteria (lines 4-7)</li>
<li>Merges the two learner results and creates html report (lines 8-9)</li>
</ul>
</div></blockquote>
<p>A Pebl configuration file can be used to create multiple learners but they must
be of the same type and use the same parameters (stopping criteria in this
case). Thus, it is not possible to replicate the above code with a
configuration file but it can be approximated:</p>
<div class="highlight-python"><table class="highlighttable"><tr><td class="linenos"><div class="linenodiv"><pre> 1
2
3
4
5
6
7
8
9
10
11
12
13</pre></div></td><td class="code"><div class="highlight"><pre><span class="p">[</span><span class="n">data</span><span class="p">]</span>
<span class="n">filename</span> <span class="o">=</span> <span class="n">pebl</span><span class="o">-</span><span class="n">tutorial</span><span class="o">-</span><span class="n">data2</span><span class="o">.</span><span class="n">txt</span>
<span class="p">[</span><span class="n">learner</span><span class="p">]</span>
<span class="nb">type</span> <span class="o">=</span> <span class="n">greedy</span><span class="o">.</span><span class="n">GreedyLearner</span>
<span class="n">numtasks</span> <span class="o">=</span> <span class="mi">2</span>
<span class="p">[</span><span class="n">greedy</span><span class="p">]</span>
<span class="n">max_iterations</span> <span class="o">=</span> <span class="mi">1000000</span>
<span class="p">[</span><span class="n">result</span><span class="p">]</span>
<span class="n">format</span> <span class="o">=</span> <span class="n">html</span>
<span class="n">outdir</span> <span class="o">=</span> <span class="n">example2</span><span class="o">-</span><span class="n">result</span>
</pre></div>
</td></tr></table></div>
</div>
<div class="section" id="third-example">
<h2>Third Example<a class="headerlink" href="#third-example" title="Permalink to this headline">¶</a></h2>
<p>For large datasets, we might wish to do multiple learner runs and use different
learners. The following example creates and runs 5 greedy and 5 simulated
annealing learners:</p>
<div class="highlight-python"><table class="highlighttable"><tr><td class="linenos"><div class="linenodiv"><pre>1
2
3
4
5
6
7</pre></div></td><td class="code"><div class="highlight"><pre><span class="gp">>>> </span><span class="kn">from</span> <span class="nn">pebl</span> <span class="kn">import</span> <span class="n">data</span><span class="p">,</span> <span class="n">result</span>
<span class="gp">>>> </span><span class="kn">from</span> <span class="nn">pebl.learner</span> <span class="kn">import</span> <span class="n">greedy</span><span class="p">,</span> <span class="n">simanneal</span>
<span class="gp">>>> </span><span class="n">dataset</span> <span class="o">=</span> <span class="n">data</span><span class="o">.</span><span class="n">fromfile</span><span class="p">(</span><span class="s">"pebl-tutorial-data2.txt"</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">learners</span> <span class="o">=</span> <span class="p">[</span> <span class="n">greedy</span><span class="o">.</span><span class="n">GreedyLearner</span><span class="p">(</span><span class="n">dataset</span><span class="p">,</span> <span class="n">max_iterations</span><span class="o">=</span><span class="mi">1000000</span><span class="p">)</span> <span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="mi">5</span><span class="p">)</span> <span class="p">]</span> <span class="o">+</span> \
<span class="gp">>>> </span> <span class="p">[</span> <span class="n">simanneal</span><span class="o">.</span><span class="n">SimulatedAnnealingLearner</span><span class="p">(</span><span class="n">dataset</span><span class="p">)</span> <span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="mi">5</span><span class="p">)</span> <span class="p">]</span>
<span class="gp">>>> </span><span class="n">merged_result</span> <span class="o">=</span> <span class="n">result</span><span class="o">.</span><span class="n">merge</span><span class="p">([</span><span class="n">learner</span><span class="o">.</span><span class="n">run</span><span class="p">()</span> <span class="k">for</span> <span class="n">learner</span> <span class="ow">in</span> <span class="n">learners</span><span class="p">])</span>
<span class="gp">>>> </span><span class="n">merged_result</span><span class="o">.</span><span class="n">tohtml</span><span class="p">(</span><span class="s">"example3-result"</span><span class="p">)</span>
</pre></div>
</td></tr></table></div>
<p>The code above is similar to the last example except that we create a list of
10 learners of two different types. The corresponding configuration file has
the same caveat as in the previous example:</p>
<div class="highlight-python"><table class="highlighttable"><tr><td class="linenos"><div class="linenodiv"><pre> 1
2
3
4
5
6
7
8
9
10
11
12
13</pre></div></td><td class="code"><div class="highlight"><pre><span class="p">[</span><span class="n">data</span><span class="p">]</span>
<span class="n">filename</span> <span class="o">=</span> <span class="n">pebl</span><span class="o">-</span><span class="n">tutorial</span><span class="o">-</span><span class="n">data2</span><span class="o">.</span><span class="n">txt</span>
<span class="p">[</span><span class="n">learner</span><span class="p">]</span>
<span class="nb">type</span> <span class="o">=</span> <span class="n">greedy</span><span class="o">.</span><span class="n">GreedyLearner</span>
<span class="n">numtasks</span> <span class="o">=</span> <span class="mi">10</span>
<span class="p">[</span><span class="n">greedy</span><span class="p">]</span>
<span class="n">max_iterations</span> <span class="o">=</span> <span class="mi">1000000</span>
<span class="p">[</span><span class="n">result</span><span class="p">]</span>
<span class="n">format</span> <span class="o">=</span> <span class="n">html</span>
<span class="n">outdir</span> <span class="o">=</span> <span class="n">example3</span><span class="o">-</span><span class="n">result</span>
</pre></div>
</td></tr></table></div>
</div>
<div class="section" id="fourth-example">
<h2>Fourth Example<a class="headerlink" href="#fourth-example" title="Permalink to this headline">¶</a></h2>
<p>In the previous example, we run 10 learners serially. We can use Pebl’s
taskcontroller package to run these learners in parallel:</p>
<div class="highlight-python"><table class="highlighttable"><tr><td class="linenos"><div class="linenodiv"><pre> 1
2
3
4
5
6
7
8
9
10</pre></div></td><td class="code"><div class="highlight"><pre><span class="gp">>>> </span><span class="kn">from</span> <span class="nn">pebl</span> <span class="kn">import</span> <span class="n">data</span><span class="p">,</span> <span class="n">result</span>
<span class="gp">>>> </span><span class="kn">from</span> <span class="nn">pebl.learner</span> <span class="kn">import</span> <span class="n">greedy</span><span class="p">,</span> <span class="n">simanneal</span>
<span class="gp">>>> </span><span class="kn">from</span> <span class="nn">pebl.taskcontroller</span> <span class="kn">import</span> <span class="n">multiprocess</span>
<span class="gp">>>> </span><span class="n">dataset</span> <span class="o">=</span> <span class="n">data</span><span class="o">.</span><span class="n">fromfile</span><span class="p">(</span><span class="s">"pebl-tutorial-data2.txt"</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">learners</span> <span class="o">=</span> <span class="p">[</span> <span class="n">greedy</span><span class="o">.</span><span class="n">GreedyLearner</span><span class="p">(</span><span class="n">dataset</span><span class="p">,</span> <span class="n">max_iterations</span><span class="o">=</span><span class="mi">1000000</span><span class="p">)</span> <span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="mi">5</span><span class="p">)</span> <span class="p">]</span> <span class="o">+</span> \
<span class="gp">>>> </span> <span class="p">[</span> <span class="n">simanneal</span><span class="o">.</span><span class="n">SimulatedAnnealingLearner</span><span class="p">(</span><span class="n">dataset</span><span class="p">)</span> <span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="mi">5</span><span class="p">)</span> <span class="p">]</span>
<span class="gp">>>> </span><span class="n">tc</span> <span class="o">=</span> <span class="n">multiprocess</span><span class="o">.</span><span class="n">MultiProcessController</span><span class="p">(</span><span class="n">poolsize</span><span class="o">=</span><span class="mi">2</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">results</span> <span class="o">=</span> <span class="n">tc</span><span class="o">.</span><span class="n">run</span><span class="p">(</span><span class="n">learners</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">merged_result</span> <span class="o">=</span> <span class="n">result</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">results</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">merged_result</span><span class="o">.</span><span class="n">tohtml</span><span class="p">(</span><span class="s">"example4-result"</span><span class="p">)</span>
</pre></div>
</td></tr></table></div>
<p>In this example, we import the multiprocess module (line 3), create a
multiprocess task controller with a pool size of two processes (line 7), run
the learners using the task controller (line 8) and merge the results and
create html report as before.</p>
<p>The corresponding configuration file (with the caveats mention in the previous
example) would be:</p>
<div class="highlight-python"><table class="highlighttable"><tr><td class="linenos"><div class="linenodiv"><pre> 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19</pre></div></td><td class="code"><div class="highlight"><pre><span class="p">[</span><span class="n">data</span><span class="p">]</span>
<span class="n">filename</span> <span class="o">=</span> <span class="n">pebl</span><span class="o">-</span><span class="n">tutorial</span><span class="o">-</span><span class="n">data1</span><span class="o">.</span><span class="n">txt</span>
<span class="p">[</span><span class="n">learner</span><span class="p">]</span>
<span class="nb">type</span> <span class="o">=</span> <span class="n">greedy</span><span class="o">.</span><span class="n">GreedyLearner</span>
<span class="n">numtasks</span> <span class="o">=</span> <span class="mi">10</span>
<span class="p">[</span><span class="n">greedy</span><span class="p">]</span>
<span class="n">max_iterations</span> <span class="o">=</span> <span class="mi">1000000</span>
<span class="p">[</span><span class="n">taskcontroller</span><span class="p">]</span>
<span class="nb">type</span> <span class="o">=</span> <span class="n">multiprocess</span><span class="o">.</span><span class="n">MultiProcessController</span>
<span class="p">[</span><span class="n">multiprocess</span><span class="p">]</span>
<span class="n">poolsize</span> <span class="o">=</span> <span class="mi">2</span>
<span class="p">[</span><span class="n">result</span><span class="p">]</span>
<span class="n">format</span> <span class="o">=</span> <span class="n">html</span>
<span class="n">outdir</span> <span class="o">=</span> <span class="n">example2</span><span class="o">-</span><span class="n">result</span>
</pre></div>
</td></tr></table></div>
<dl class="docutils">
<dt>Pebl provides three other task controllers:</dt>
<dd><ul class="first last simple">
<li><a class="reference internal" href="taskcontroller/xgrid.html#module-pebl.taskcontroller.xgrid" title="pebl.taskcontroller.xgrid: Xgrid task controller"><tt class="xref py py-mod docutils literal"><span class="pre">pebl.taskcontroller.xgrid</span></tt></a> for using Apple’s XGrid</li>
<li><a class="reference internal" href="taskcontroller/ipy1.html#module-pebl.taskcontroller.ipy1" title="pebl.taskcontroller.ipy1: IPython1 task controller"><tt class="xref py py-mod docutils literal"><span class="pre">pebl.taskcontroller.ipy1</span></tt></a> for using an Ipython1 cluster</li>
<li><a class="reference internal" href="taskcontroller/ec2.html#module-pebl.taskcontroller.ec2" title="pebl.taskcontroller.ec2: Amazon EC2 task controller"><tt class="xref py py-mod docutils literal"><span class="pre">pebl.taskcontroller.ec2</span></tt></a> for using Amazon EC2</li>
</ul>
</dd>
</dl>
<p>All task controllers can be used with the pebl application and configuration
file and the only difference between their usage are the parameters they
require. Thus, Pebl allows one to do preliminary analysis on their desktop with
perhaps the multiprocess task controller and then do the full analysis using an
XGrid or Amazon’s EC2 by simply changing one line of code or a few lines in a
configuration file. The EC2 task controller is an especially attractive option
for large analysis tasks because it allows one to rent the computing resources
on an as-needed basis and without any cluster installation or configuration.</p>
</div>
<div class="section" id="a-note-on-interpreting-the-results">
<h2>A Note on Interpreting the Results<a class="headerlink" href="#a-note-on-interpreting-the-results" title="Permalink to this headline">¶</a></h2>
<p>There is no principled way to determine the optimal stopping criteria or
simulated annealing parameters for analyzing a given dataset. One common
strategy is to construct consensus networks that show network features found
with high confidence. Pebl’s html reports show such “model-averaged” networks in
the third tab and the <a class="reference internal" href="posterior.html#module-pebl.posterior" title="pebl.posterior: Posterior distribution"><tt class="xref py py-mod docutils literal"><span class="pre">pebl.posterior</span></tt></a> module has methods for creating
these programatically.</p>
<p>Another common strategy is to check for stability of results. You begin with
some learning, save the results, do futher learning, merge the two results and
see if the top networks and consensus networks have changed much. If they
remain relatively stable, you can assume that you’ve reached a good solution.
Keep in mind, however, that you can never guarantee that you’ve found the
optimal network (or that there is a singular optimal network to be found) since
structure learning of Bayesian network is a known NP-Hard problem.</p>
<p>In the examples above, we’ve been creating html reports of the results but these
cannot be later merged. A better option is to save the result using the
<a class="reference internal" href="result.html#pebl.result.LearnerResult.tofile" title="pebl.result.LearnerResult.tofile"><tt class="xref py py-meth docutils literal"><span class="pre">pebl.result.LearnerResult.tofile()</span></tt></a> method and then later read it with
<a class="reference internal" href="result.html#pebl.result.fromfile" title="pebl.result.fromfile"><tt class="xref py py-func docutils literal"><span class="pre">pebl.result.fromfile()</span></tt></a>:</p>
<div class="highlight-python"><table class="highlighttable"><tr><td class="linenos"><div class="linenodiv"><pre>1
2
3
4
5
6
7
8</pre></div></td><td class="code"><div class="highlight"><pre><span class="gp">>>> </span><span class="kn">from</span> <span class="nn">pebl</span> <span class="kn">import</span> <span class="n">data</span><span class="p">,</span> <span class="n">result</span>
<span class="gp">>>> </span><span class="n">result1</span> <span class="o">=</span> <span class="n">learner</span><span class="o">.</span><span class="n">run</span><span class="p">()</span>
<span class="gp">>>> </span><span class="n">result1</span><span class="o">.</span><span class="n">tofile</span><span class="p">(</span><span class="s">"result1.pebl"</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">result1</span><span class="o">.</span><span class="n">tohtml</span><span class="p">(</span><span class="s">"result1"</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">result2</span> <span class="o">=</span> <span class="n">otherlearner</span><span class="o">.</span><span class="n">run</span><span class="p">()</span>
<span class="gp">>>> </span><span class="n">result1</span> <span class="o">=</span> <span class="n">result</span><span class="o">.</span><span class="n">fromfile</span><span class="p">(</span><span class="s">"result1.pebl"</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">merged_result</span> <span class="o">=</span> <span class="n">result</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">result1</span><span class="p">,</span> <span class="n">result2</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">merged_result</span><span class="o">.</span><span class="n">tofile</span><span class="p">(</span><span class="s">"results_sofar.pebl"</span><span class="p">)</span>
</pre></div>
</td></tr></table></div>
<p>A third strategy is to calculate a p-value for each scored network. This will be added to the tutorial shortly.</p>
</div>
<div class="section" id="a-note-on-scale-of-problems-pebl-can-solve">
<h2>A Note on Scale of Problems Pebl Can Solve<a class="headerlink" href="#a-note-on-scale-of-problems-pebl-can-solve" title="Permalink to this headline">¶</a></h2>
<p>Pebl imposes few limits on the scale of problems it can handle other than the
limits imposed by the hardware you use. The
<a class="reference internal" href="paramref.html#confparam-localscore_cache.maxsize"><tt class="xref std std-confparam docutils literal"><span class="pre">localscore_cache.maxsize</span></tt></a> configuration parameter can be used to
control the size of the main cache used by pebl. With a value appropriate to
your memory availability, pebl can be used for quite large datasets. We have
successfully tested pebl with datasets of size (number variable, number
samples) = (10000,10000) and (1000,100000) on a machine with 2GB memory.</p>
<p>While pebl can handle such large datasets without crashing, because structure
learning is a known NP-Hard problem, using pebl with datasets containing more
than a few hundred variables will likely give poor results due to poor coverage
of the search space.</p>
</div>
<div class="section" id="more-coming-soon">
<h2>More Coming Soon<a class="headerlink" href="#more-coming-soon" title="Permalink to this headline">¶</a></h2>
<p>I will be adding examples for using prior knowledge and for calculating p-values using a bootstrapping approach.</p>
</div>
<div class="section" id="learning-more">
<h2>Learning More<a class="headerlink" href="#learning-more" title="Permalink to this headline">¶</a></h2>
<p>This tutorial should have given you an overview of using Pebl. For further
information about specific components, consult the <a class="reference internal" href="apiref.html#apiref"><em>API Reference</em></a>, which
contains detailed information about all parts of pebl. If you would like to
add code to pebl, consult the <a class="reference internal" href="devguide.html#devguide"><em>Developer’s Guide</em></a>. Feel free to contact me (Abhik
Shah <<a class="reference external" href="mailto:abhikshah%40gmail.com">abhikshah<span>@</span>gmail<span>.</span>com</a>>) with any questions or comments.</p>
</div>
<div class="section" id="bibliography">
<h2>Bibliography<a class="headerlink" href="#bibliography" title="Permalink to this headline">¶</a></h2>
<dl class="docutils">
<dt>[1] Spellman et al., (1998). Comprehensive Identification of Cell</dt>
<dd>Cycle-regulated Genes of the Yeast Saccharomyces cerevisiae by Microarray
Hybridization. Molecular Biology of the Cell 9, 3273-3297.</dd>
<dt>[2] Husmeier et al., Probabilistic Modeling in Bioinformatics and Medical</dt>
<dd>Informatics. Springer, 2004. <a class="reference external" href="http://books.google.com/books?id=ND8rjHNkJ-QC">http://books.google.com/books?id=ND8rjHNkJ-QC</a></dd>
</dl>
</div>
</div>
</div>
</div>
</div>
<div class="sphinxsidebar">
<div class="sphinxsidebarwrapper">
<h3><a href="index.html">Table Of Contents</a></h3>
<ul>
<li><a class="reference internal" href="#">Tutorial</a><ul>
<li><a class="reference internal" href="#introducing-the-problem">Introducing the Problem</a></li>
<li><a class="reference internal" href="#first-example">First Example</a></li>
<li><a class="reference internal" href="#pebl-s-data-file-format">Pebl’s Data File Format</a></li>
<li><a class="reference internal" href="#second-example">Second Example</a></li>
<li><a class="reference internal" href="#third-example">Third Example</a></li>
<li><a class="reference internal" href="#fourth-example">Fourth Example</a></li>
<li><a class="reference internal" href="#a-note-on-interpreting-the-results">A Note on Interpreting the Results</a></li>
<li><a class="reference internal" href="#a-note-on-scale-of-problems-pebl-can-solve">A Note on Scale of Problems Pebl Can Solve</a></li>
<li><a class="reference internal" href="#more-coming-soon">More Coming Soon</a></li>
<li><a class="reference internal" href="#learning-more">Learning More</a></li>
<li><a class="reference internal" href="#bibliography">Bibliography</a></li>
</ul>
</li>
</ul>
<h4>Previous topic</h4>
<p class="topless"><a href="install.html"
title="previous chapter">Installation</a></p>
<h4>Next topic</h4>
<p class="topless"><a href="devguide.html"
title="next chapter">Developer’s Guide</a></p>
<h3>This Page</h3>
<ul class="this-page-menu">
<li><a href="_sources/tutorial.txt"
rel="nofollow">Show Source</a></li>
</ul>
<div id="searchbox" style="display: none">
<h3>Quick search</h3>
<form class="search" action="search.html" method="get">
<input type="text" name="q" size="18" />
<input type="submit" value="Go" />
<input type="hidden" name="check_keywords" value="yes" />
<input type="hidden" name="area" value="default" />
</form>
<p class="searchtip" style="font-size: 90%">
Enter search terms or a module, class or function name.
</p>
</div>
<script type="text/javascript">$('#searchbox').show(0);</script>
</div>
</div>
<div class="clearer"></div>
</div>
<div class="related">
<h3>Navigation</h3>
<ul>
<li class="right" style="margin-right: 10px">
<a href="genindex.html" title="General Index"
>index</a></li>
<li class="right" >
<a href="py-modindex.html" title="Python Module Index"
>modules</a> |</li>
<li class="right" >
<a href="devguide.html" title="Developer’s Guide"
>next</a> |</li>
<li class="right" >
<a href="install.html" title="Installation"
>previous</a> |</li>
<li><a href="index.html">Pebl v1.0.1 documentation</a> »</li>
</ul>
</div>
<div class="footer">
© Copyright 2008, Abhik Shah.
Last updated on Apr 29, 2011.
Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.0.7.
</div>
</body>
</html>
|