-
Notifications
You must be signed in to change notification settings - Fork 1
/
how-many-copies.html
392 lines (327 loc) · 32.3 KB
/
how-many-copies.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
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
<!DOCTYPE html>
<html lang="en" prefix="og: http://ogp.me/ns# fb: https://www.facebook.com/2008/fbml">
<head>
<title>How many copies - Stackdiver as a Service</title>
<!-- Using the latest rendering mode for IE -->
<meta http-equiv="X-UA-Compatible" content="IE=edge">
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<link rel="canonical" href="https://www.lyh.me/how-many-copies.html">
<meta name="author" content="Neville Li" />
<meta name="keywords" content="data,performance,scala" />
<meta name="description" content="One topic that came up a lot when optimizing Scala data applications is the performance of standard collections, or the hidden cost of temporary copies. The collections API is easy to learn and maps well to many Python concepts where a lot of data engineers are familiar with. But the performance penalty can be pretty big when it’s repeated over millions of records in a JVM with limited heap. Mapping values Let’s take a look at one most naive example first, mapping the values of a Map. val m = Map("A" -> 1, "B" -> 2, "C" -> 3) m.toList.map(t => (t._1, t._2 + 1)).toMap Looks simple enough but obviously not optimal. Two temporary List[(String, Int)] were created, one from toList and one from map. map also creates 3 copies of (String, Int). There are a few commonly seen variations. These don’t create temporary collections but still key-value tuples. for ((k, v) <- m) yield k -> (v + 1) m.map { case (k, v) => k -> (v + 1) } If one reads the ScalaDoc closely, there’s a mapValues method already and it probably is the shortest and most performant. m.mapValues(_ + 1) Java conversion Similar problem exists …" />
<meta property="og:site_name" content="Stackdiver as a Service" />
<meta property="og:type" content="article"/>
<meta property="og:title" content="How many copies"/>
<meta property="og:url" content="https://www.lyh.me/how-many-copies.html"/>
<meta property="og:description" content="One topic that came up a lot when optimizing Scala data applications is the performance of standard collections, or the hidden cost of temporary copies. The collections API is easy to learn and maps well to many Python concepts where a lot of data engineers are familiar with. But the performance penalty can be pretty big when it’s repeated over millions of records in a JVM with limited heap. Mapping values Let’s take a look at one most naive example first, mapping the values of a Map. val m = Map("A" -> 1, "B" -> 2, "C" -> 3) m.toList.map(t => (t._1, t._2 + 1)).toMap Looks simple enough but obviously not optimal. Two temporary List[(String, Int)] were created, one from toList and one from map. map also creates 3 copies of (String, Int). There are a few commonly seen variations. These don’t create temporary collections but still key-value tuples. for ((k, v) <- m) yield k -> (v + 1) m.map { case (k, v) => k -> (v + 1) } If one reads the ScalaDoc closely, there’s a mapValues method already and it probably is the shortest and most performant. m.mapValues(_ + 1) Java conversion Similar problem exists …"/>
<meta property="article:published_time" content="2014-08-02" />
<meta property="article:section" content="code" />
<meta property="article:tag" content="data" />
<meta property="article:tag" content="performance" />
<meta property="article:tag" content="scala" />
<meta property="article:author" content="Neville Li" />
<!-- Bootstrap -->
<link rel="stylesheet" href="https://www.lyh.me/theme/css/bootstrap.min.css" type="text/css"/>
<link href="https://www.lyh.me/theme/css/font-awesome.min.css" rel="stylesheet">
<link href="https://www.lyh.me/theme/css/pygments/monokai.css" rel="stylesheet">
<link href="https://www.lyh.me/theme/css/typogrify.css" rel="stylesheet">
<link rel="stylesheet" href="https://www.lyh.me/theme/css/style.css" type="text/css"/>
<link href="https://www.lyh.me/feeds/all.atom.xml" type="application/atom+xml" rel="alternate"
title="Stackdiver as a Service ATOM Feed"/>
<link href="https://www.lyh.me/feeds/code.atom.xml" type="application/atom+xml" rel="alternate"
title="Stackdiver as a Service code ATOM Feed"/>
</head>
<body>
<div class="navbar navbar-default navbar-fixed-top" role="navigation">
<div class="container">
<div class="navbar-header">
<button type="button" class="navbar-toggle" data-toggle="collapse" data-target=".navbar-ex1-collapse">
<span class="sr-only">Toggle navigation</span>
<span class="icon-bar"></span>
<span class="icon-bar"></span>
<span class="icon-bar"></span>
</button>
<a href="https://www.lyh.me/" class="navbar-brand">
Stackdiver as a Service </a>
</div>
<div class="collapse navbar-collapse navbar-ex1-collapse">
<ul class="nav navbar-nav">
<li><a href="https://www.lyh.me/pages/about-me.html">
About Me
</a></li>
</ul>
<ul class="nav navbar-nav navbar-right">
</ul>
</div>
<!-- /.navbar-collapse -->
</div>
</div> <!-- /.navbar -->
<!-- Banner -->
<!-- End Banner -->
<!-- Content Container -->
<div class="container">
<div class="row">
<div class="col-sm-9">
<section id="content">
<article>
<header class="page-header">
<h1>
<a href="https://www.lyh.me/how-many-copies.html"
rel="bookmark"
title="Permalink to How many copies">
How many copies
</a>
</h1>
</header>
<div class="entry-content">
<div class="panel">
<div class="panel-body">
<footer class="post-info">
<span class="label label-default">Date</span>
<span class="published">
<i class="fa fa-calendar"></i><time datetime="2014-08-02T20:48:00-04:00"> Sat 02 August 2014</time>
</span>
<span class="label label-default">Category</span>
<a href="https://www.lyh.me/category/code.html">code</a>
<span class="label label-default">Tags</span>
<a href="https://www.lyh.me/tag/data.html">data</a>
/
<a href="https://www.lyh.me/tag/performance.html">performance</a>
/
<a href="https://www.lyh.me/tag/scala.html">scala</a>
</footer><!-- /.post-info --> </div>
</div>
<p>One topic that came up a lot when optimizing Scala data applications is the performance of standard collections, or the hidden cost of temporary copies. The collections <span class="caps">API</span> is easy to learn and maps well to many Python concepts where a lot of data engineers are familiar with. But the performance penalty can be pretty big when it’s repeated over millions of records in a <span class="caps">JVM</span> with limited heap.</p>
<h2>Mapping values</h2>
<p>Let’s take a look at one most naive example first, mapping the values of a <code>Map</code>.</p>
<div class="highlight"><pre><span></span><code><span class="k">val</span> <span class="n">m</span> <span class="k">=</span> <span class="nc">Map</span><span class="o">(</span><span class="s">"A"</span> <span class="o">-></span> <span class="mi">1</span><span class="o">,</span> <span class="s">"B"</span> <span class="o">-></span> <span class="mi">2</span><span class="o">,</span> <span class="s">"C"</span> <span class="o">-></span> <span class="mi">3</span><span class="o">)</span>
<span class="n">m</span><span class="o">.</span><span class="n">toList</span><span class="o">.</span><span class="n">map</span><span class="o">(</span><span class="n">t</span> <span class="k">=></span> <span class="o">(</span><span class="n">t</span><span class="o">.</span><span class="n">_1</span><span class="o">,</span> <span class="n">t</span><span class="o">.</span><span class="n">_2</span> <span class="o">+</span> <span class="mi">1</span><span class="o">)).</span><span class="n">toMap</span>
</code></pre></div>
<p>Looks simple enough but obviously not optimal. Two temporary <code>List[(String, Int)]</code> were created, one from <code>toList</code> and one from <code>map</code>. <code>map</code> also creates 3 copies of <code>(String, Int)</code>.</p>
<p>There are a few commonly seen variations. These don’t create temporary collections but still key-value tuples.</p>
<div class="highlight"><pre><span></span><code><span class="k">for</span> <span class="o">((</span><span class="n">k</span><span class="o">,</span> <span class="n">v</span><span class="o">)</span> <span class="k"><-</span> <span class="n">m</span><span class="o">)</span> <span class="k">yield</span> <span class="n">k</span> <span class="o">-></span> <span class="o">(</span><span class="n">v</span> <span class="o">+</span> <span class="mi">1</span><span class="o">)</span>
<span class="n">m</span><span class="o">.</span><span class="n">map</span> <span class="o">{</span> <span class="k">case</span> <span class="o">(</span><span class="n">k</span><span class="o">,</span> <span class="n">v</span><span class="o">)</span> <span class="k">=></span> <span class="n">k</span> <span class="o">-></span> <span class="o">(</span><span class="n">v</span> <span class="o">+</span> <span class="mi">1</span><span class="o">)</span> <span class="o">}</span>
</code></pre></div>
<p>If one reads the <a href="http://www.scala-lang.org/api/2.10.4/index.html#scala.collection.immutable.Map">ScalaDoc</a> closely, there’s a <code>mapValues</code> method already and it probably is the shortest and most performant.</p>
<div class="highlight"><pre><span></span><code><span class="n">m</span><span class="o">.</span><span class="n">mapValues</span><span class="o">(</span><span class="k">_</span> <span class="o">+</span> <span class="mi">1</span><span class="o">)</span>
</code></pre></div>
<h2>Java conversion</h2>
<p>Similar problem exists when converting between Scala and Java collections. We store data on <span class="caps">HDFS</span> in <a href="http://avro.apache.org/">Avro</a> and one most commonly used data structure is high dimensional vector, represented as <code>java.util.List[java.lang.Float]</code> in Avro and <code>Array[Float]</code> in Scala and <a href="https://github.com/scalanlp/breeze">breeze</a>.</p>
<p>Since there’s no direct conversion between Java <code>List</code> and Scala <code>Array</code>, the naive solution requires a temporary copy and looks like this:</p>
<div class="highlight"><pre><span></span><code><span class="k">import</span> <span class="nn">java.util.</span><span class="o">{</span> <span class="nc">List</span> <span class="k">=></span> <span class="nc">JList</span> <span class="o">}</span>
<span class="k">import</span> <span class="nn">java.lang.</span><span class="o">{</span> <span class="nc">Float</span> <span class="k">=></span> <span class="nc">JFloat</span> <span class="o">}</span>
<span class="k">import</span> <span class="nn">scala.collection.JavaConverters._</span>
<span class="k">def</span> <span class="n">javaToArray</span><span class="o">(</span><span class="n">v</span><span class="k">:</span> <span class="kt">JList</span><span class="o">[</span><span class="kt">JFloat</span><span class="o">])</span><span class="k">:</span> <span class="kt">Array</span><span class="o">[</span><span class="kt">Float</span><span class="o">]</span> <span class="k">=</span> <span class="n">v</span><span class="o">.</span><span class="n">asScala</span><span class="o">.</span><span class="n">asInstanceOf</span><span class="o">[</span><span class="kt">List</span><span class="o">[</span><span class="kt">Float</span><span class="o">]].</span><span class="n">toArray</span>
<span class="k">def</span> <span class="n">arrayToJava</span><span class="o">(</span><span class="n">v</span><span class="k">:</span> <span class="kt">Array</span><span class="o">[</span><span class="kt">Float</span><span class="o">])</span><span class="k">:</span> <span class="kt">JList</span><span class="o">[</span><span class="kt">JFloat</span><span class="o">]</span> <span class="k">=</span> <span class="n">v</span><span class="o">.</span><span class="n">toList</span><span class="o">.</span><span class="n">asInstanceOf</span><span class="o">[</span><span class="kt">List</span><span class="o">[</span><span class="kt">JFloat</span><span class="o">]].</span><span class="n">asJava</span>
</code></pre></div>
<p>The conversion is almost 10 times faster by doing the conversion manually in a loop.</p>
<div class="highlight"><pre><span></span><code><span class="k">def</span> <span class="n">javaToArray</span><span class="o">(</span><span class="n">v</span><span class="k">:</span> <span class="kt">JList</span><span class="o">[</span><span class="kt">JFloat</span><span class="o">])</span><span class="k">:</span> <span class="kt">Array</span><span class="o">[</span><span class="kt">Float</span><span class="o">]</span> <span class="k">=</span> <span class="o">{</span>
<span class="k">val</span> <span class="n">r</span> <span class="k">=</span> <span class="k">new</span> <span class="nc">Array</span><span class="o">[</span><span class="kt">Float</span><span class="o">](</span><span class="n">v</span><span class="o">.</span><span class="n">size</span><span class="o">)</span>
<span class="k">var</span> <span class="n">i</span> <span class="k">=</span> <span class="mi">0</span>
<span class="k">val</span> <span class="n">j</span> <span class="k">=</span> <span class="n">v</span><span class="o">.</span><span class="n">iterator</span>
<span class="k">while</span> <span class="o">(</span><span class="n">i</span> <span class="o"><</span> <span class="n">v</span><span class="o">.</span><span class="n">size</span><span class="o">)</span> <span class="o">{</span>
<span class="n">r</span><span class="o">(</span><span class="n">i</span><span class="o">)</span> <span class="k">=</span> <span class="n">j</span><span class="o">.</span><span class="n">next</span>
<span class="n">i</span> <span class="o">+=</span> <span class="mi">1</span>
<span class="o">}</span>
<span class="n">r</span>
<span class="o">}</span>
<span class="k">def</span> <span class="n">arrayToJava</span><span class="o">(</span><span class="n">v</span><span class="k">:</span> <span class="kt">Array</span><span class="o">[</span><span class="kt">Float</span><span class="o">])</span><span class="k">:</span> <span class="kt">JList</span><span class="o">[</span><span class="kt">JFloat</span><span class="o">]</span> <span class="k">=</span> <span class="o">{</span>
<span class="k">val</span> <span class="n">r</span> <span class="k">=</span> <span class="k">new</span> <span class="nc">JArrayList</span><span class="o">[</span><span class="kt">JFloat</span><span class="o">](</span><span class="n">v</span><span class="o">.</span><span class="n">size</span><span class="o">)</span>
<span class="k">var</span> <span class="n">i</span> <span class="k">=</span> <span class="mi">0</span>
<span class="k">while</span> <span class="o">(</span><span class="n">i</span> <span class="o"><</span> <span class="n">v</span><span class="o">.</span><span class="n">size</span><span class="o">)</span> <span class="o">{</span>
<span class="n">r</span><span class="o">.</span><span class="n">add</span><span class="o">(</span><span class="n">v</span><span class="o">(</span><span class="n">i</span><span class="o">))</span>
<span class="n">i</span> <span class="o">+=</span> <span class="mi">1</span>
<span class="o">}</span>
<span class="n">r</span>
<span class="o">}</span>
</code></pre></div>
<h2>Merging maps</h2>
<p>Same goes for merging 2 <code>Map[String, Double]</code> by summing up values of duplicate keys. Both solutions below require 2 copies of key sets, one union/intersect set,</p>
<div class="highlight"><pre><span></span><code><span class="k">val</span> <span class="n">m1</span> <span class="k">=</span> <span class="nc">Map</span><span class="o">(</span><span class="s">"A"</span> <span class="o">-></span> <span class="mf">1.0</span><span class="o">,</span> <span class="s">"B"</span> <span class="o">-></span> <span class="mf">2.0</span><span class="o">,</span> <span class="s">"C"</span> <span class="o">-></span> <span class="mf">3.0</span><span class="o">)</span>
<span class="k">val</span> <span class="n">m2</span> <span class="k">=</span> <span class="nc">Map</span><span class="o">(</span><span class="s">"A"</span> <span class="o">-></span> <span class="mf">1.5</span><span class="o">,</span> <span class="s">"B"</span> <span class="o">-></span> <span class="mf">2.5</span><span class="o">,</span> <span class="s">"D"</span> <span class="o">-></span> <span class="mf">3.5</span><span class="o">)</span>
<span class="c1">// solution 1</span>
<span class="k">val</span> <span class="n">i</span> <span class="k">=</span> <span class="n">m1</span><span class="o">.</span><span class="n">keySet</span> <span class="n">intersect</span> <span class="n">m2</span><span class="o">.</span><span class="n">keySet</span>
<span class="k">val</span> <span class="n">m</span> <span class="k">=</span> <span class="n">i</span><span class="o">.</span><span class="n">map</span> <span class="o">{</span> <span class="n">k</span> <span class="k">=></span> <span class="n">k</span> <span class="o">-></span> <span class="o">(</span><span class="n">m1</span><span class="o">(</span><span class="n">k</span><span class="o">)</span> <span class="o">+</span> <span class="n">m2</span><span class="o">(</span><span class="n">k</span><span class="o">))</span> <span class="o">}</span>
<span class="o">(</span><span class="n">m1</span> <span class="o">--</span> <span class="n">i</span><span class="o">)</span> <span class="o">++</span> <span class="o">(</span><span class="n">m2</span> <span class="o">--</span> <span class="n">i</span><span class="o">)</span> <span class="o">++</span> <span class="n">m</span> <span class="c1">// -- i unnecessary</span>
<span class="n">m1</span> <span class="o">++</span> <span class="n">m2</span> <span class="o">++</span> <span class="n">m</span> <span class="c1">// slightly better</span>
<span class="c1">// solution 2</span>
<span class="o">(</span><span class="n">m1</span><span class="o">.</span><span class="n">keySet</span> <span class="o">++</span> <span class="n">m2</span><span class="o">.</span><span class="n">keySet</span><span class="o">)</span> <span class="n">map</span> <span class="o">{</span> <span class="n">k</span> <span class="k">=></span>
<span class="n">k</span> <span class="o">-></span> <span class="o">(</span><span class="n">m1</span><span class="o">.</span><span class="n">getOrElse</span><span class="o">(</span><span class="n">k</span><span class="o">,</span> <span class="mf">0.0</span><span class="o">)</span> <span class="o">+</span> <span class="n">m2</span><span class="o">.</span><span class="n">getOrElse</span><span class="o">(</span><span class="n">k</span><span class="o">,</span> <span class="mf">0.0</span><span class="o">))</span>
<span class="o">}</span>
</code></pre></div>
<p>The best solution for this case also happens to be the shortest. It leverages the fact that pairs from the right-hand size of <code>++</code> overwrite those on the left-hand size. One could even compare size of the maps and put the smaller on the right-hand size to further reduce copies.</p>
<div class="highlight"><pre><span></span><code><span class="n">m1</span> <span class="o">++</span> <span class="n">m2</span><span class="o">.</span><span class="n">map</span> <span class="o">{</span> <span class="k">case</span> <span class="o">(</span><span class="n">k</span><span class="o">,</span> <span class="n">v</span><span class="o">)</span> <span class="k">=></span> <span class="n">k</span> <span class="o">-></span> <span class="o">(</span><span class="n">v</span> <span class="o">+</span> <span class="n">m1</span><span class="o">.</span><span class="n">getOrElse</span><span class="o">(</span><span class="n">k</span><span class="o">,</span> <span class="mf">0.0</span><span class="o">))</span> <span class="o">}</span>
</code></pre></div>
<h2>Cheating by mutation</h2>
<p>All of these optimizations are nice but sometimes they are just not enough. Imagine if the map merge operation is used in a <code>foldLeft</code> on the reducer size, over millions of users with one <code>Map</code> each, we’ll be basically creating that many temporary copies and sometimes <span class="caps">GC</span> just can’t keep up. Good thing is that in some cases we know for sure a temporary object won’t be reused ever again, and can mutate it in place and pass it along to further reduce copying. In the example below <span class="caps">LHS</span> is mutated in place and returned in the binary operation.</p>
<div class="highlight"><pre><span></span><code><span class="k">import</span> <span class="nn">scala.collection.mutable.</span> <span class="o">{</span> <span class="nc">Map</span> <span class="k">=></span> <span class="nc">MMap</span> <span class="o">}</span>
<span class="k">def</span> <span class="n">mergeMaps</span><span class="o">(</span><span class="n">m1</span><span class="k">:</span> <span class="kt">MMap</span><span class="o">[</span><span class="kt">String</span><span class="p">,</span> <span class="kt">Double</span><span class="o">],</span> <span class="n">m2</span><span class="k">:</span> <span class="kt">MMap</span><span class="o">[</span><span class="kt">String</span><span class="p">,</span> <span class="kt">Double</span><span class="o">])</span> <span class="k">=</span> <span class="o">{</span>
<span class="n">m2</span><span class="o">.</span><span class="n">foreach</span> <span class="o">{</span> <span class="k">case</span> <span class="o">(</span><span class="n">k</span><span class="o">,</span> <span class="n">v</span><span class="o">)</span> <span class="k">=></span> <span class="n">m1</span><span class="o">(</span><span class="n">k</span><span class="o">)</span> <span class="k">=</span> <span class="n">v</span> <span class="o">+</span> <span class="n">m1</span><span class="o">.</span><span class="n">getOrElse</span><span class="o">(</span><span class="n">k</span><span class="o">,</span> <span class="mf">0.0</span><span class="o">)</span> <span class="o">}</span>
<span class="n">m1</span>
<span class="o">}</span>
</code></pre></div>
<p>And in this <code>foldLeft</code> operation, <span class="caps">LHS</span> is always the same object as initial value <code>z</code> and mutated throughout the group.</p>
<div class="highlight"><pre><span></span><code><span class="k">val</span> <span class="n">group</span><span class="k">:</span> <span class="kt">Grouped</span><span class="o">[</span><span class="kt">String</span><span class="p">,</span> <span class="kt">MMap</span><span class="o">[</span><span class="kt">String</span><span class="p">,</span> <span class="kt">Double</span><span class="o">]]</span> <span class="c1">// = ...</span>
<span class="n">group</span><span class="o">.</span><span class="n">foldLeft</span><span class="o">(</span><span class="nc">MMap</span><span class="o">[</span><span class="kt">String</span><span class="p">,</span> <span class="kt">Double</span><span class="o">]())(</span><span class="n">mergeMaps</span><span class="o">)</span>
</code></pre></div>
<p>Some numeric libraries, like <a href="http://mikiobraun.github.io/jblas/"><span class="caps">JBLAS</span></a> offer operators in both immutable (by copying) and mutable (by mutating in place) fashions. It’s probably best to start with the immutable ones to ensure correctness, and drop down to mutating in place when running into performance problems.</p>
</div>
<!-- /.entry-content -->
<hr />
<!-- AddThis Button BEGIN -->
<div class="addthis_toolbox addthis_default_style">
<a class="addthis_button_facebook_like" fb:like:layout="button_count"></a>
<a class="addthis_button_tweet"></a>
<a class="addthis_button_google_plusone" g:plusone:size="medium"></a>
</div>
<!-- AddThis Button END -->
<hr/>
<section class="comments" id="comments">
<h2>Comments</h2>
<div id="disqus_thread"></div>
<script type="text/javascript">
/* * * CONFIGURATION VARIABLES: EDIT BEFORE PASTING INTO YOUR WEBPAGE * * */
var disqus_shortname = 'lyh'; // required: replace example with your forum shortname
var disqus_config = function () {
this.language = "en";
this.page.identifier = '2014-08-02-how-many-copies';
this.page.url = 'https://www.lyh.me/how-many-copies.html';
};
/* * * DON'T EDIT BELOW THIS LINE * * */
(function () {
var dsq = document.createElement('script');
dsq.type = 'text/javascript';
dsq.async = true;
dsq.src = '//' + disqus_shortname + '.disqus.com/embed.js';
(document.getElementsByTagName('head')[0] || document.getElementsByTagName('body')[0]).appendChild(dsq);
})();
</script>
<noscript>Please enable JavaScript to view the <a href="http://disqus.com/?ref_noscript">comments powered by
Disqus.</a></noscript>
<a href="http://disqus.com" class="dsq-brlink">comments powered by <span class="logo-disqus">Disqus</span></a>
</section>
</article>
</section>
</div>
<div class="col-sm-3" id="sidebar">
<aside>
<div id="aboutme">
<p>
<img width="100%" class="img-thumbnail" src="https://www.lyh.me//avatar.jpg"/>
</p>
<p>
<strong>About Neville Li</strong><br/>
Data infrastructure @<a href="https://twitter.com/Spotify">Spotify</a>, ex-@<a href="https://twitter.com/Yahoo">Yahoo</a> search, creator of <a href="https://github.com/spotify/scio">Scio</a>, technical cave & wreck diver, lefty guitar player
</p>
</div><!-- Sidebar -->
<section class="well well-sm">
<ul class="list-group list-group-flush">
<!-- Sidebar/Social -->
<li class="list-group-item">
<h4><i class="fa fa-home fa-lg"></i><span class="icon-label">Social</span></h4>
<ul class="list-group" id="social">
<li class="list-group-item"><a href="https://open.spotify.com/user/sinisa_lyh"><i class="fa fa-spotify fa-lg"></i> Spotify</a></li>
<li class="list-group-item"><a href="https://github.com/nevillelyh"><i class="fa fa-github-square fa-lg"></i> GitHub</a></li>
<li class="list-group-item"><a href="https://twitter.com/sinisa_lyh"><i class="fa fa-twitter-square fa-lg"></i> Twitter</a></li>
<li class="list-group-item"><a href="https://www.slideshare.net/sinisalyh"><i class="fa fa-slideshare fa-lg"></i> SlideShare</a></li>
<li class="list-group-item"><a href="https://www.youtube.com/user/sinisalyh/videos"><i class="fa fa-youtube-square fa-lg"></i> YouTube</a></li>
<li class="list-group-item"><a href="https://www.instagram.com/sinisa/"><i class="fa fa-instagram fa-lg"></i> Instagram</a></li>
<li class="list-group-item"><a href="https://www.flickr.com/photos/sinisa_lyh"><i class="fa fa-flickr fa-lg"></i> Flickr</a></li>
</ul>
</li>
<!-- End Sidebar/Social -->
<!-- Sidebar/Recent Posts -->
<li class="list-group-item">
<h4><i class="fa fa-home fa-lg"></i><span class="icon-label">Recent Posts</span></h4>
<ul class="list-group" id="recentposts">
<li class="list-group-item"><a href="https://www.lyh.me/magnolify.html">Magnolify</a></li>
<li class="list-group-item"><a href="https://www.lyh.me/featran.html">Featran</a></li>
<li class="list-group-item"><a href="https://www.lyh.me/automatic-type-class-derivation-with-shapeless.html">Automatic type-class derivation with Shapeless</a></li>
<li class="list-group-item"><a href="https://www.lyh.me/lambda-serialization.html">Lambda serialization</a></li>
<li class="list-group-item"><a href="https://www.lyh.me/lawfulness-of-aggregatebykey.html">Lawfulness of aggregateByKey</a></li>
</ul>
</li>
<!-- End Sidebar/Recent Posts -->
<!-- Sidebar/Categories -->
<li class="list-group-item">
<h4><i class="fa fa-home fa-lg"></i><span class="icon-label">Categories</span></h4>
<ul class="list-group" id="categories">
<li class="list-group-item">
<a href="https://www.lyh.me/category/code.html"><i class="fa fa-folder-open fa-lg"></i>code</a>
</li>
<li class="list-group-item">
<a href="https://www.lyh.me/category/misc.html"><i class="fa fa-folder-open fa-lg"></i>misc</a>
</li>
</ul>
</li>
<!-- End Sidebar/Categories -->
<!-- Sidebar/Twitter Timeline -->
<li class="list-group-item">
<h4><i class="fa fa-twitter fa-lg"></i><span class="icon-label">Latest Tweets</span></h4>
<div id="twitter_timeline">
<a class="twitter-timeline" data-width="250" data-height="300" data-dnt="true" data-theme="light" href="https://twitter.com/sinisa_lyh">Tweets by sinisa_lyh</a> <script async src="//platform.twitter.com/widgets.js" charset="utf-8"></script>
</div>
</li>
<!-- End Sidebar/Twitter Timeline -->
</ul>
</section>
<!-- End Sidebar --> </aside>
</div>
</div>
</div>
<!-- End Content Container -->
<footer>
<div class="container">
<hr>
<div class="row">
<div class="col-xs-10">© 2020 Neville Li
· Powered by <a href="https://github.com/getpelican/pelican-themes/tree/master/pelican-bootstrap3" target="_blank">pelican-bootstrap3</a>,
<a href="http://docs.getpelican.com/" target="_blank">Pelican</a>,
<a href="http://getbootstrap.com" target="_blank">Bootstrap</a> <p><small> <a rel="license" href="https://creativecommons.org/licenses/by-nc/4.0/deed.en"><img alt="Creative Commons License" style="border-width:0" src="//i.creativecommons.org/l/by-nc/4.0/80x15.png" /></a>
Content
licensed under a <a rel="license" href="https://creativecommons.org/licenses/by-nc/4.0/deed.en">Creative Commons Attribution-NonCommercial 4.0 International License</a>, except where indicated otherwise.
</small></p>
</div>
<div class="col-xs-2"><p class="pull-right"><i class="fa fa-arrow-up"></i> <a href="#">Back to top</a></p></div>
</div>
</div>
</footer>
<script src="https://www.lyh.me/theme/js/jquery.min.js"></script>
<!-- Include all compiled plugins (below), or include individual files as needed -->
<script src="https://www.lyh.me/theme/js/bootstrap.min.js"></script>
<!-- Enable responsive features in IE8 with Respond.js (https://github.com/scottjehl/Respond) -->
<script src="https://www.lyh.me/theme/js/respond.min.js"></script>
<!-- Disqus -->
<script type="text/javascript">
/* * * CONFIGURATION VARIABLES: EDIT BEFORE PASTING INTO YOUR WEBPAGE * * */
var disqus_shortname = 'lyh'; // required: replace example with your forum shortname
/* * * DON'T EDIT BELOW THIS LINE * * */
(function () {
var s = document.createElement('script');
s.async = true;
s.type = 'text/javascript';
s.src = '//' + disqus_shortname + '.disqus.com/count.js';
(document.getElementsByTagName('HEAD')[0] || document.getElementsByTagName('BODY')[0]).appendChild(s);
}());
</script>
<!-- End Disqus Code -->
<!-- Google Analytics -->
<script type="text/javascript">
var _gaq = _gaq || [];
_gaq.push(['_setAccount', 'UA-6988688-5']);
_gaq.push(['_trackPageview']);
(function () {
var ga = document.createElement('script');
ga.type = 'text/javascript';
ga.async = true;
ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js';
var s = document.getElementsByTagName('script')[0];
s.parentNode.insertBefore(ga, s);
})();
</script>
<!-- End Google Analytics Code -->
<script type="text/javascript">var addthis_config = {"data_track_addressbar": true};</script>
<script type="text/javascript" src="//s7.addthis.com/js/300/addthis_widget.js#pubid=sinisalyh"></script>
</body>
</html>