/
index.html
720 lines (572 loc) · 25.1 KB
/
index.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
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
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<title>Python @Mapbox</title>
<meta name="description" content="A framework for easily creating beautiful presentations using HTML">
<meta name="author" content="Hakim El Hattab">
<meta name="apple-mobile-web-app-capable" content="yes" />
<meta name="apple-mobile-web-app-status-bar-style" content="black-translucent" />
<meta name="viewport" content="width=device-width, initial-scale=1.0, maximum-scale=1.0, user-scalable=no">
<link rel="stylesheet" href="css/reveal.min.css">
<link rel="stylesheet" href="css/theme/default.css" id="theme">
<!-- For syntax highlighting -->
<link rel="stylesheet" href="lib/css/zenburn.css">
<!-- If the query includes 'print-pdf', include the PDF print sheet -->
<script>
if( window.location.search.match( /print-pdf/gi ) ) {
var link = document.createElement( 'link' );
link.rel = 'stylesheet';
link.type = 'text/css';
link.href = 'css/print/pdf.css';
document.getElementsByTagName( 'head' )[0].appendChild( link );
}
</script>
<!--[if lt IE 9]>
<script src="lib/js/html5shiv.js"></script>
<![endif]-->
</head>
<body>
<div class="reveal">
<div class="slides">
<section
data-background="img/rgba30.png"
data-background-size="1000px">
<h1>Python at Mapbox</h1>
<p>Sean Gillies</p>
<p><small>Fort Collins Python Web Development Meetup∙July 14, 2014</small></p>
<aside class="notes">
Hi! Good evening! Thank you very much for coming out to the Meetup.
My name
is Sean Gillies, I work at Mapbox.
I've been in the GIS business since 1999 and I've been a Python programmer
since 2001.
</aside>
</section>
<section
data-background="img/3PmjkxZ.gif"
data-background-size="1024px">
<aside class="notes">
Mapbox is a platform for creating and using maps. That means that it’s
a collection of tools and services that connect together in different ways:
you can draw a map on the website, use it on an iPhone, and get raw data
from an API.
We quite frankly want to win over the users of the Google Maps platform.
This evening I'm going to cover:
- How Mapbox works
- How Mapbox employees work
- What the satellite team is up to
- How we're using Python
So, starting with a look at the products and the company, I'm going to
bend back towards Python :)
</aside>
</section>
<section>
<h2>How Mapbox works</h2>
<ul>
<li class="fragment">Making maps</li>
<li class="fragment">Map data</li>
<li class="fragment">Using maps</li>
</ul>
</section>
<section>
<h2>Making maps: overlays</h2>
<img src="img/drawing-polygon.png">
<aside class="notes">
Fundamentally, maps are made up of overlays and baselayers. Overlays are additional data added on top of a map for some particular purpose. An overlay might be a marker with your business location, or something complex like earthquake visualizations. These sit on top of the baselayer - a combination of streets, countries, and terrain or satellite data that give people a basic sense of place.
</aside>
</section>
<section>
<h2>Making maps: baselayers</h2>
<img src="img/all_layers.jpg">
<aside class="notes">
The easiest way to make a map is with the Mapbox.com editor. You can customize parameters of the baselayer - choosing satellite or terrain, adjusting road or area colors, or changing the language of labels. It also provides a simple interface to draw, style, and import features like markers, lines, and areas.
The Mapbox.com editor is great for getting started fast, and has some superpowers over any other way of making maps. With it, you can tweak the baselayer, and instantly, everywhere in the world acquires a new look, on the fly. But sometimes you want more control, or to add more data, and for that, there’s TileMill.
</aside>
</section>
<section>
<h2>Making maps: tilemill</h2>
<img width="640" src="img/project.png">
<aside class="notes">
TileMill is a free desktop application that can import and style data to your heart’s content. It can help you create everything from heatmaps to 3D buildings and custom terrain. You can even export the features you drew on Mapbox.com and import them into TileMill for more styling power. And while the Mapbox.com editor tops out at 2,000 features and might get slow with complex geometries, TileMill stays super fast.
Generating big maps in TileMill? You might want to check out our upgraded plans for more storage space.
The TileMill approach is not without limitations - unlike the Mapbox.com editor, you can’t instantly style a map of the world. If you’re dealing with a large geographical area and high zoom levels, it can take a while to render and will result in big files.
So TileMill is a desktop application - how does it connect to Mapbox? Well, we’ve made it so that it’s easy to upload straight from TileMill to Mapbox and integrate your TileMill maps with your Mapbox.com projects.
</aside>
</section>
<section>
<h2>Mapbox map data</h2>
<ul>
<li>Streets: OpenStreetMap</li>
<li>Imagery: DigitalGlobe and other providers</li>
<li>Common to all customers</li>
</ul>
<aside class="notes">
Let’s look at Mapbox’s data first. The baselayer that you customize in Mapbox.com comes from a variety of sources: the road network comes from OpenStreetMap, while satellite and terrain data are derived from DigitalGlobe and other image providers. We carefully choose and curate our sources to make sure that everything from country borders to the Sahara is accurate.
</aside>
</section>
<section>
<img src="img/tCDSKJX.gif">
<aside class="notes">
This week, the data team finished adding buildings in San Francisco to OpenStreetMap. Since January, we have added or updated over 152,022 building polygons based on San Francisco building footprint data and Bing imagery.
The biggest difference between Google Maps and Mapbox isn't size. It's that
Mapbox uses and gives back to OSM.
We're big contributors to OSM.
Edits made to OSM show up in Mapbox data in less than 10 minutes.
</aside>
</section>
<section>
<h2>Added map data</h2>
<img src="img/markers-3.gif">
<aside class="notes">
Next, there’s data you add in Mapbox.com or stylize in TileMill. This can be any compatible source you choose - and there are several compatible formats for TileMill as well as for the editor. This should be data that you own or have a right to use. We make sure that you can always download it - either by exporting it from the web interface or keeping it on your local computer.
</aside>
</section>
<section>
<h2>Local map data</h2>
<ul>
<li>Not uploaded to Mapbox</li>
<li>Combined with maps in realtime</li>
<li>Using API or SDK</li>
</ul>
<aside class="notes">
There’s a third kind of data: local data. Local data can be combined with your maps in realtime, thanks to the power of our SDKs. This can be anything from dynamic database results to files that users upload. Unlike other kinds of data, this never has to touch Mapbox’s servers because it’s combined with the map in realtime. A good example of this kind of integration is Foursquare, a company that uses Mapbox’s custom maps and overlays their own custom data with custom software.
</aside>
</section>
<section>
<h2>Using maps on the web</h2>
<img src="img/firstmap-sharelink.png">
<aside class="notes">
The simplest way of sharing a map is just sending a link: this directs someone to Mapbox.com where they can look at your map through a nice little interface. Additionally, you can use embeds, which let you add a map within your existing content, like you would add an image to a blog post.
</aside>
</section>
<section>
<h2>JS, you say?</h2>
<p>Mapbox.js is a great plugin for Leaflet.</p>
<aside class="notes">
And there are a bunch of related plugins for loading and mapping
all kinds of files: CSV, KML, GPX.
BTW, Leaflet's author works at Mapbox.
</aside>
</section>
<section>
<h2>Using maps on mobile</h2>
<img style="border: none" src="img/mobile.png">
<aside class="notes">
Native devices call for their own code: so there’s the Mapbox iOS SDK on iOS, MBXMapKit for iOS 7+, and the Mapbox Android SDK for any flavor of Android, including AOSP. Just like Mapbox.js, we build all of these tools as free and open source projects: you can customize them, peek inside, and try them out.
Regardless of how you access Mapbox, whether through an SDK, embeds, or using the core API, you’re building on the Mapbox cloud service, so your custom maps will be available and pixel-perfect.
</aside>
</section>
<section>
<img src="img/8466078950_a66740cb32_k.jpg">
<aside class="notes">
I'm now going to explain how we work at Mapbox.
It's like a lot of other offices: pizza, cola, and satellite launches.
That's Landsat 8.
From the Washington DC office. We also have an office in SF.
</aside>
</section>
<section>
<h2>Distributed is how we work</h2>
<ul>
<li>30 in Washington, DC</li>
<li>15 in San Francisco</li>
<li>12 remote: US, Canada, Europe, South America</li>
</ul>
<aside class="notes">
Including Fort Collins and, very recently, Boulder.
Mapbox's funders, The Foundry Group, are in Boulder, too.
</aside>
</section>
<section>
<h2>Distributed is really how we work</h2>
<ul>
<li>GitHub for everything</li>
<li class="fragment">for code</li>
<li class="fragment">for supplies</li>
<li class="fragment">for travel planning</li>
<li class="fragment">for business strategy</li>
</ul>
<aside class="notes">
It feels a lot like working on a bunch of open source projects.
Everybody has at least a couple job descriptions and we rotate around.
Now, flat and open isn't a panacea. In fact, it can be a recipe for abuse
if you're hiring people purely on their ambition and tech pedigrees. So
we try not to be irrationally enthusiastic about lack of managers and
mertitocracy.
We're also trying to equalize the gender ratio at Mapbox. Being a woman
in tech should be normal, we think.
</aside>
</section>
<section>
<img src="img/14606867651_49cfefde0e_b.jpg">
<aside class="notes">
We do lots of social stuff, too.
We host meetups, hackathons, mapping parties,
sponsor local women's orgs.
We totally enjoy this.
</aside>
</section>
<section>
<img src="img/14467410326_966200aa48_k.jpg">
<aside class="notes">
Trolling the establishment is also a small part of what we do at
Mapbox.
</aside>
</section>
<!-- Cloudless Gibralter -->
<section
data-background="img/8871945777_603bfd7d9f_b.jpg"
data-background-size="1024px">
<h2>Mapbox Cloudless Atlas</h2>
<aside class="notes">
The Mapbox Satellite Team is making a Cloudless Landsat mosaic of the
world.
This is a base layer for maps created by Mapbox users. Its primary
requirements are to be current and to look good. We'll probably be working
on products that are better for science in the future.
This is the project that drives a lot of Rasterio development.
It has interesting technical problems.
</aside>
</section>
<!-- Scene sources -->
<section
data-background="img/8871364724_5d48143feb_o.jpg"
data-background-size="1024px">
<h2>From cloudy scenes</h2>
<aside class="notes">
We're making the cloudless atlas from generally cloudy scenes.
The first technical problem is getting the clouds out.
Everywhere.
With minimal operator intervention.
</section>
<!-- Scene sources -->
<section>
<h2>Declouding</h2>
<ul>
<li>Scene selection (season and cloudiness)</li>
<li>Atmospheric correction and reflectance normalization</li>
<li>Score every pixel of every source image</li>
<li class="fragment">No cutlines</li>
<li class="fragment">Pixel by pixel</li>
</ul>
<aside class="notes">
Cutlines and other spatial interventions don't scale well.
Cloudless Atlas needs to scale, because we want to run it almost
continually as new data comes in.
Pixel by pixel approaches scale trivially.
They worked very well for MODIS imagery. They won't work for very high
resolution imagery? Will they work for Landsat? At the outset we weren't
sure.
</aside>
</section>
<!-- good pixels sorted up -->
<section
data-background="img/8870292339_517612af6d_k.jpg"
data-background-size="1024px"
data-background-position="top">
<h2>High scoring pixels</h2>
<aside class="notes">
This is our approach right now: score every pixel and sort the best
ones to the top. The cloudless atlas tiles are formed from the best
pixels.
Shown here are the "best" by some measure (darkest in this case) pixels
of all the Landsat scenes shown in the previous slide.
</aside>
</section>
<!-- bad pixels sorted down -->
<section
data-background="img/8870292339_517612af6d_k.jpg"
data-background-size="1024px"
data-background-position="bottom">
<h2>Low scoring pixels</h2>
<aside class="notes">
These are the worst (brightest) pixels.
Sorting is fairly cheap (no worse than NlogN - N being the number of
input scenes), so this is going to scale
reasonably well.
</aside>
</section>
<section>
<h2>Earth is a cloudy planet</h2>
<ul>
<li>50-70% mean cloud cover over land</li>
<li>Landsat 8 doesn't yet have cloud-free imagery everywhere</li>
<li class="fragment">So we also use Landsat 7</li>
</ul>
<aside class="notes">
This is the lead in to our second technical problem. We have to use
Landsat 7 because we don't yet have enough L8 coverage to make a
global cloudless mosaic. And Landsat 7 has some problems.
</section>
<section>
<h2>Scan Line Correction</h2>
<img src="img/products_slc_graphic_s.jpg" width="200%">
<aside class="notes">
Landsat 7's scan line corrector, which accounts for the forward motion of
the satellite as it scans, failed permanently on May 31, 2003. Imagery
acquired after this time has zig-zaging lines which both overlap and miss
portions of the landscape.
</aside>
</section>
<section
data-background="img/8871307188_29f0c482f6_b.jpg"
data-background-size="1024px">
<aside class="notes">
SLC-Off imagery has characteristic no-data strips, increasing in
breadth toward the edges
of the scene. Middles of the scenes can have full coverage.
</aside>
</section>
<section>
<h2>Landsat SLC-Off Imagery</h2>
<ul>
<li>Since May 2003</li>
<li>Without it we don't have enough cloudless pixels</li>
<li class="fragment">It complicates things</li>
<li class="fragment">A lot</li>
</ul>
<aside class="notes">
Many regions are cloudy enough that the best pixels are found only in
SLC-Off scenes.
Putting SLC-Off imagery into our declouding pipeline leads to our
second interesting technical problem.
</section>
<section
data-background="img/8871945763_17ff3e37f4_b.jpg"
data-background-size="1024px">
<h2>Bands</h2>
<aside class="notes">
Instead of nodata bands we have bands of seasonal or interannual
incongruity.
It can be
especially noticeable in agricultural regions: manifesting as
alternating strips of vigorous and non-vigorous vegetation (or bare
ground).
</section>
<section>
<h2>Debanding</h2>
<ul>
<li class="fragment">Transform the image using a Fast Fourier Transform</li>
<li class="fragment">Locate the artifact peaks</li>
<li class="fragment">Remove the artifacts</li>
<li class="fragment">Transform the image back using an inverse transform</li>
</ul>
<aside class="notes">
This is our approach to the debanding problem.
Locating the artifact peaks is quite difficult in practice when there are
multiple SLC-off images going in to our product.
Modeling the orientation of the SLC-off gaps helps a lot.
</aside>
</section>
<!--section>
<h2>Debanding (2)</h2>
<ul>
<li>Uses numpy.fft, scikit-image, etc.</li>
<li>And orbital information</li>
<li>Multiple iterations needed to prevent addition of image bugs</li>
<li>Details: <a href="https://www.mapbox.com/blog/debanding-the-world/">https://www.mapbox.com/blog/debanding-the-world/</a></li>
</ul>
<aside class="notes">
Modeling the orientation of SLC-off bands lets us focus on just those particular
peaks.
</aside>
</section-->
<section>
<h2>Solving the problems</h2>
<ul>
<li>We have no off-the-shelf solutions</li>
<li>Trial and error is involved</li>
</ul>
<aside class="notes">
Trial and error isn't our only method, but it's an important one. It
has a big impact on our requirements.
</aside>
</section>
<section>
<h2>Cloudless Atlas Software Requirements</h2>
<ol>
<li class="fragment">Rapid prototyping and iteration</li>
<li class="fragment">Proven algorithms and drivers</li>
<li class="fragment">Ability to deploy to 100s of servers</li>
</ol>
<aside class="notes">
We need to get new iterations of our algorithms into production
quickly. We need robust numerical libraries. And we need to distribute
the work across many servers. And also know the cost of doing it.
</section>
<section>
<h2>Software stack</h2>
<ul>
<li>1 and 3 argue for open source</li>
<li>1 argues for a high-level language with handy multi-dimensional array syntax</li>
<li>2 argues for LAPACK (&c) and GDAL</li>
<li class="fragment">The fit: Linux, Python, Numpy, Scipy</li>
</ul>
</section>
<!-- Geospatial -->
<section>
<h2>GDAL Python bindings?</h2>
<ul>
<li>Very C++ style interface</li>
<li>Failure to keep SWIG and C issues in mind results in crashing programs</li>
<li class="fragment">We can do better</li>
</ul>
<aside class="notes">
The last thing the stack needs is software to read and write geospatial
raster data files like GeoTIFFs.
And reproject them.
GDAL has Python bindings but they are not IMHO production grade.
They have terrible gotchas.
They also feel too much like C++. That's no fun. programming at a higher
level and high productivity are our requirements, remember?
</aside>
</section>
<section>
<h2>Scipy-style raster data library requirements</h2>
<ol>
<li>Read/write ndarrays from/to data files</li>
<li>Python types, protocols and idioms instead of C/C++ ones</li>
<li>Free programmers from having to think about C/C++ (memory management, pointers, &c)</li>
</ol>
</section>
<section>
<h2>Rasterio's Design</h2>
<ul>
<li>A Python package at the top</li>
<li>Extension modules (using Cython) in the middle</li>
<li>Fast loops, typed memoryviews, "nogil" blocks</li>
<li>GDAL shared library on the bottom</li>
</ul>
<aside class="notes">
Cython takes care of a lot of C extension details for us.
Rasterio uses Cython features like fast loops, typed memoryviews,
and "nogil" blocks.
</section>
<section>
<h2>Reading data</h2>
<pre><code data-trim>
import rasterio
with rasterio.open(path) as src:
bands = [src.read_band(i) for i in src.indexes]
</code></pre>
<ul><li><code>open()</code> gives you a file-like dataset object</li>
<li><code>read_band()</code> gives you a Numpy ndarray</li>
<li>Read windows of data with extended slice-like syntax</li></ul>
<aside class="notes">
Rasterio is geared for band-interleaved data.
New read() method coming in 0.9.
</aside>
</section>
<section>
<h2>Writing data</h2>
<pre><code data-trim>
kwargs = src.meta
with rasterio.open(path, 'w', **kwargs) as dst:
for i, arr in zip(dst.indexes, bands):
dst.write_band(i, arr)
</code></pre>
<ul><li>Get keyword args needed to open a dataset for writing from another dataset</li>
<li><code>write_band()</code> takes an ndarray</li>
<li>You can also write to windows of a dataset</li></ul>
<aside class="notes">
Rasterio is geared for band-interleaved data.
New read() method coming in 0.9.
</aside>
</section>
<section>
<h2>Georeferencing</h2>
Rasterio follows the lead of pyproj
<pre><code data-trim>
>>> import rasterio
>>> src = rasterio.open('rasterio/tests/data/RGB.byte.tif')
>>> src.crs
{'units': 'm', 'zone': 18, 'ellps': 'WGS84', 'proj': 'utm', 'no_defs': True}
</code></pre>
</section>
<section>
<h2>rasterio.features</h2>
<ul><li><code>rasterio.features.shapes()</code> yields all the features of an array as GeoJSON-like objects</li>
<li><code>rasterio.features.rasterize()</code> "burns" GeoJSON-like objects into an array</li>
<li>Dicts, iterators, tuples, ndarrays</li>
<li>No datasets or layers necessary</li></ul>
</section>
<section>
<img src="img/py-powered.png">
<pre><code data-trim>
{'coordinates': [[(71.0, 6.0), ...]], 'type': 'Polygon'}, ...
</code></pre>
<img src="img/py-powered-rev.png">
</section>
<section>
<h2>rasterio.warp</h2>
<ul><li><code>rasterio.warp.reproject()</code> maps elements of one array to another, using cartographic projections</li>
<li>No datasets or layers required</li>
<li>Data created in non-GIS programs can be reprojected for use with GIS programs</li></ul>
</section>
<section>
<!-- What I told you -->
<h2>Conclusion</h2>
<ul>
<li class="fragment">The Mapbox Satellite team has interesting problems</li>
<li class="fragment">Linux, Python, Numpy, Scipy, GDAL, and Rasterio are big parts of the solutions</li>
<li class="fragment">Rasterio aims to make GIS data more accessible to Python programmers</li>
<li class="fragment">And help GIS analysts learn important Python protocols and idioms</li>
</ul>
<aside class="notes">
In conclusion.
</aside>
</section>
<section
data-background="img/rgba30.png"
data-background-size="1000px">
<h1>Rasterio</h1>
<a href="https://github.com/mapbox/rasterio">https://github.com/mapbox/rasterio</a>
</section>
<section>
<h2>Debanding Blog Post</h2>
<a href="https://www.mapbox.com/blog/debanding-the-world/">https://www.mapbox.com/blog/debanding-the-world/</a> by Amit Kapadia
</section>
<section
data-background="img/rgba30.png"
data-background-size="1000px">
<h2>Thank You!</h2>
<p><a href="https://twitter.com/sgillies">@sgillies</a></p>
<p><a href="mailto:sean@mapbox.com">sean@mapbox.com</a></p>
</section>
<section>
<h2>THE END</h2>
</section>
<!-- end slides -->
</div>
<!-- end reveal container -->
</div>
<!-- Scripts below -->
<script src="lib/js/head.min.js"></script>
<script src="js/reveal.min.js"></script>
<script>
// Full list of configuration options available here:
// https://github.com/hakimel/reveal.js#configuration
Reveal.initialize({
controls: true,
progress: true,
history: true,
center: true,
theme: Reveal.getQueryHash().theme, // available themes are in /css/theme
transition: Reveal.getQueryHash().transition || 'default', // default/cube/page/concave/zoom/linear/fade/none
// Parallax scrolling
// parallaxBackgroundImage: 'https://s3.amazonaws.com/hakim-static/reveal-js/reveal-parallax-1.jpg',
// parallaxBackgroundSize: '2100px 900px',
// Optional libraries used to extend on reveal.js
dependencies: [
{ src: 'lib/js/classList.js', condition: function() { return !document.body.classList; } },
{ src: 'plugin/markdown/marked.js', condition: function() { return !!document.querySelector( '[data-markdown]' ); } },
{ src: 'plugin/markdown/markdown.js', condition: function() { return !!document.querySelector( '[data-markdown]' ); } },
{ src: 'plugin/highlight/highlight.js', async: true, callback: function() { hljs.initHighlightingOnLoad(); } },
{ src: 'plugin/zoom-js/zoom.js', async: true, condition: function() { return !!document.body.classList; } },
{ src: 'plugin/notes/notes.js', async: true, condition: function() { return !!document.body.classList; } }
]
});
</script>
</body>
</html>