-
Notifications
You must be signed in to change notification settings - Fork 6
/
Motivation.html
336 lines (271 loc) · 13 KB
/
Motivation.html
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
<!DOCTYPE html>
<html>
<head>
<title>Should I use HPC?</title>
<meta charset="utf-8">
<meta http-equiv="X-UA-Compatible" content="chrome=1">
<meta name="generator" content="pandoc" />
<meta name="viewport" content="width=device-width, initial-scale=1">
<meta name="apple-mobile-web-app-capable" content="yes">
<base target="_blank">
<script type="text/javascript">
var SLIDE_CONFIG = {
// Slide settings
settings: {
title: 'Should I use HPC?',
useBuilds: true,
usePrettify: true,
enableSlideAreas: true,
enableTouch: true,
favIcon: 'Motivation_files/logo.png',
},
// Author information
presenters: [
{
name: 'Pedro Ojeda' ,
company: '',
gplus: '',
twitter: '',
www: '',
github: ''
},
]
};
</script>
<script src="site_libs/header-attrs-2.7/header-attrs.js"></script>
<link href="site_libs/ioslides-13.5.1/fonts/fonts.css" rel="stylesheet" />
<link href="site_libs/ioslides-13.5.1/theme/css/default.css" rel="stylesheet" />
<link href="site_libs/ioslides-13.5.1/theme/css/phone.css" rel="stylesheet" />
<script src="site_libs/ioslides-13.5.1/js/modernizr.custom.45394.js"></script>
<script src="site_libs/ioslides-13.5.1/js/prettify/prettify.js"></script>
<script src="site_libs/ioslides-13.5.1/js/prettify/lang-r.js"></script>
<script src="site_libs/ioslides-13.5.1/js/prettify/lang-yaml.js"></script>
<script src="site_libs/ioslides-13.5.1/js/hammer.js"></script>
<script src="site_libs/ioslides-13.5.1/js/slide-controller.js"></script>
<script src="site_libs/ioslides-13.5.1/js/slide-deck.js"></script>
<style type="text/css">
b, strong {
font-weight: bold;
}
em {
font-style: italic;
}
summary {
display: list-item;
}
slides > slide {
-webkit-transition: all 0.4s ease-in-out;
-moz-transition: all 0.4s ease-in-out;
-o-transition: all 0.4s ease-in-out;
transition: all 0.4s ease-in-out;
}
.auto-fadein {
-webkit-transition: opacity 0.6s ease-in;
-webkit-transition-delay: 0.4s;
-moz-transition: opacity 0.6s ease-in 0.4s;
-o-transition: opacity 0.6s ease-in 0.4s;
transition: opacity 0.6s ease-in 0.4s;
opacity: 0;
}
/* https://github.com/ropensci/plotly/pull/524#issuecomment-468142578 */
slide:not(.current) .plotly.html-widget{
display: block;
}
code{white-space: pre-wrap;}
span.smallcaps{font-variant: small-caps;}
span.underline{text-decoration: underline;}
div.column{display: inline-block; vertical-align: top; width: 50%;}
div.hanging-indent{margin-left: 1.5em; text-indent: -1.5em;}
ul.task-list{list-style: none;}
pre > code.sourceCode { white-space: pre; position: relative; }
pre > code.sourceCode > span { display: inline-block; line-height: 1.25; }
pre > code.sourceCode > span:empty { height: 1.2em; }
code.sourceCode > span { color: inherit; text-decoration: inherit; }
div.sourceCode { margin: 1em 0; }
pre.sourceCode { margin: 0; }
@media screen {
div.sourceCode { overflow: auto; }
}
@media print {
pre > code.sourceCode { white-space: pre-wrap; }
pre > code.sourceCode > span { text-indent: -5em; padding-left: 5em; }
}
pre.numberSource code
{ counter-reset: source-line 0; }
pre.numberSource code > span
{ position: relative; left: -4em; counter-increment: source-line; }
pre.numberSource code > span > a:first-child::before
{ content: counter(source-line);
position: relative; left: -1em; text-align: right; vertical-align: baseline;
border: none; display: inline-block;
-webkit-touch-callout: none; -webkit-user-select: none;
-khtml-user-select: none; -moz-user-select: none;
-ms-user-select: none; user-select: none;
padding: 0 4px; width: 4em;
color: #aaaaaa;
}
pre.numberSource { margin-left: 3em; border-left: 1px solid #aaaaaa; padding-left: 4px; }
div.sourceCode
{ }
@media screen {
pre > code.sourceCode > span > a:first-child::before { text-decoration: underline; }
}
code span.al { color: #ff0000; font-weight: bold; } /* Alert */
code span.an { color: #60a0b0; font-weight: bold; font-style: italic; } /* Annotation */
code span.at { color: #7d9029; } /* Attribute */
code span.bn { color: #40a070; } /* BaseN */
code span.bu { } /* BuiltIn */
code span.cf { color: #007020; font-weight: bold; } /* ControlFlow */
code span.ch { color: #4070a0; } /* Char */
code span.cn { color: #880000; } /* Constant */
code span.co { color: #60a0b0; font-style: italic; } /* Comment */
code span.cv { color: #60a0b0; font-weight: bold; font-style: italic; } /* CommentVar */
code span.do { color: #ba2121; font-style: italic; } /* Documentation */
code span.dt { color: #902000; } /* DataType */
code span.dv { color: #40a070; } /* DecVal */
code span.er { color: #ff0000; font-weight: bold; } /* Error */
code span.ex { } /* Extension */
code span.fl { color: #40a070; } /* Float */
code span.fu { color: #06287e; } /* Function */
code span.im { } /* Import */
code span.in { color: #60a0b0; font-weight: bold; font-style: italic; } /* Information */
code span.kw { color: #007020; font-weight: bold; } /* Keyword */
code span.op { color: #666666; } /* Operator */
code span.ot { color: #007020; } /* Other */
code span.pp { color: #bc7a00; } /* Preprocessor */
code span.sc { color: #4070a0; } /* SpecialChar */
code span.ss { color: #bb6688; } /* SpecialString */
code span.st { color: #4070a0; } /* String */
code span.va { color: #19177c; } /* Variable */
code span.vs { color: #4070a0; } /* VerbatimString */
code span.wa { color: #60a0b0; font-weight: bold; font-style: italic; } /* Warning */
slides > slide:not(.nobackground):before {
font-size: 12pt;
content: "";
position: absolute;
bottom: 20px;
left: 60px;
background: url(Motivation_files/logo.png) no-repeat 0 50%;
-webkit-background-size: 30px 30px;
-moz-background-size: 30px 30px;
-o-background-size: 30px 30px;
background-size: 30px 30px;
padding-left: 40px;
height: 30px;
line-height: 1.9;
}
</style>
<link rel="stylesheet" href="custom.css" type="text/css" />
</head>
<body style="opacity: 0">
<slides class="layout-widescreen">
<slide class="title-slide segue nobackground">
<aside class="gdbar"><img src="Motivation_files/logo.png"></aside>
<!-- The content of this hgroup is replaced programmatically through the slide_config.json. -->
<hgroup class="auto-fadein">
<h1 data-config-title><!-- populated from slide_config.json --></h1>
<p data-config-presenter><!-- populated from slide_config.json --></p>
<p style="margin-top: 6px; margin-left: -2px;">Feb., 2021</p>
</hgroup>
</slide>
<slide class=""><hgroup><h2>Iris data set</h2></hgroup><article id="iris-data-set">
<p>We will use the <em>iris</em> database:</p>
<pre class = 'prettyprint lang-r'>x <- iris[which(iris[,5] != "setosa"), c(1,5)]
summary(x)</pre>
<pre >## Sepal.Length Species
## Min. :4.900 setosa : 0
## 1st Qu.:5.800 versicolor:50
## Median :6.300 virginica :50
## Mean :6.262
## 3rd Qu.:6.700
## Max. :7.900</pre>
<p>Example adapted from: <a href='https://unc-libraries-data.github.io/R-Open-Labs/Extras/Parallel/foreach.html' title=''>https://unc-libraries-data.github.io/R-Open-Labs/Extras/Parallel/foreach.html</a></p>
</article></slide><slide class=""><hgroup><h2>Iris data set</h2></hgroup><article id="iris-data-set-1">
<pre class = 'prettyprint lang-r'>plot(x[,1],x[,2]) #plot of Species vs. sepal lengths</pre>
<p><img src="Motivation_files/figure-html/unnamed-chunk-2-1.png" width="720" /></p>
</article></slide><slide class=""><hgroup><h2>Serial mode</h2></hgroup><article id="serial-mode">
<p>we will use the <em>foreach</em> function from the <em>doParallel</em> package to fit the data to a generalized linear model.</p>
<pre class = 'prettyprint lang-r'>library(doParallel)</pre>
</article></slide><slide class=""><hgroup><h2>Serial mode</h2></hgroup><article id="serial-mode-1">
<p>Let’s take a look at the performance of a logistic regression model in serial mode (1 core):</p>
<pre class = 'prettyprint lang-r'>stime <- system.time({
r <- foreach(1:10000, .combine=cbind) %do% {
train <- sample(100,100, replace=TRUE)
result1 <- glm(x[train,2]~x[train,1], family=binomial(logit))
coefficients(result1)
}
})[3]
stime</pre>
<pre >## elapsed
## 22.24</pre>
</article></slide><slide class=""><hgroup><h2>Parallel mode</h2></hgroup><article id="parallel-mode">
<p>Now, look at the performance using 2 cores:</p>
<pre class = 'prettyprint lang-r'>cl <- makeCluster(2)
registerDoParallel(cl)
ptime <- system.time({
r <- foreach(1:10000, .combine=cbind) %dopar% {
train <- sample(100,100, replace=TRUE)
result1 <- glm(x[train,2]~x[train,1], family=binomial(logit))
coefficients(result1)
}
})[3]
ptime</pre>
<pre >## elapsed
## 16.11</pre>
<pre class = 'prettyprint lang-r'>stopCluster(cl)</pre>
</article></slide><slide class=""><hgroup><h2>Parallel mode</h2></hgroup><article id="parallel-mode-1">
<p>a graphical view of the scaling behavior can be seen in the following plot:</p>
<p><img src="Motivation_files/figure-html/unnamed-chunk-6-1.png" width="720" /></p>
</article></slide><slide class=""><hgroup><h2>Is parallel processing always the best alternative?</h2></hgroup><article id="is-parallel-processing-always-the-best-alternative">
<pre class = 'prettyprint lang-r'>stime <- system.time(
foreach(i=1:1e4) %do% sqrt(i) )
stime </pre>
<pre >## user system elapsed
## 1.73 0.00 1.73</pre>
<pre class = 'prettyprint lang-r'>cl <- makeCluster(2)
registerDoParallel(cl)
ptime <- system.time(
foreach(i=1:1e4) %dopar% sqrt(i) )
ptime </pre>
<pre >## user system elapsed
## 2.65 0.26 3.24</pre>
<pre class = 'prettyprint lang-r'>stopCluster(cl)</pre>
</article></slide><slide class=""><hgroup><h2>Is parallel processing always the best alternative?</h2></hgroup><article id="is-parallel-processing-always-the-best-alternative-1">
<p>Only if the computational load (number of numerical operations) exceeds the overhead of using the parallel routines.</p>
<p>Another message from these examples is that one always needs to check the performance of the code (Time) vs. the number of requested cores, to use an “optimal” number.</p>
<p>These data can be useful upon applying for medium/large SNIC projects because the reviewers would know if the requested hours are justified.</p>
</article></slide><slide class=""><hgroup><h2>References</h2></hgroup><article id="references">
<ul>
<li><a href='https://swcarpentry.github.io/r-novice-inflammation/' title=''>https://swcarpentry.github.io/r-novice-inflammation/</a></li>
<li><a href='https://www.tutorialspoint.com/r/index.htm' title=''>https://www.tutorialspoint.com/r/index.htm</a></li>
<li>R High Performance Programming. Aloysius, Lim; William, Tjhi. Packt Publishing, 2015.</li>
<li><a href='https://www.r-bloggers.com/the-wonders-of-foreach/' title=''>https://www.r-bloggers.com/the-wonders-of-foreach/</a></li>
<li><a href='https://unc-libraries-data.github.io/R-Open-Labs/Extras/Parallel/foreach.html' title=''>https://unc-libraries-data.github.io/R-Open-Labs/Extras/Parallel/foreach.html</a></li>
</ul>
<p><a href='index.html' title=''>Return to Index</a></p></article></slide>
<slide class="backdrop"></slide>
</slides>
<!-- dynamically load mathjax for compatibility with self-contained -->
<script>
(function () {
var script = document.createElement("script");
script.type = "text/javascript";
script.src = "https://mathjax.rstudio.com/latest/MathJax.js?config=TeX-AMS-MML_HTMLorMML";
document.getElementsByTagName("head")[0].appendChild(script);
})();
</script>
<!-- map slide visiblity events into shiny -->
<script>
(function() {
if (window.jQuery) {
window.jQuery(document).on('slideleave', function(e) {
window.jQuery(e.target).trigger('hidden');
});
window.jQuery(document).on('slideenter', function(e) {
window.jQuery(e.target).trigger('shown');
});
}
})();
</script>
</body>
</html>