PyCon 2013

https://us.pycon.org/2013/

관련자료
 * [[http://pyvideo.org/category/33/pycon-us-2013|Videos]]
 * [[https://speakerdeck.com/pyconslides/|Slides]]
 * [[yong27]] 포스팅 [[http://yong27.biohackers.net/s2/5/pycon-2013/|PyCon 2013에 다녀와서]]

|| <<TableOfContents>> ||

= Tutorials =
== Fast Python Program through Optimization ==
https://us.pycon.org/2013/schedule/presentation/24/

http://www.python-academy.com Mile Muller
 * handout: http://www.python-academy.com/download/pycon2013/optimization_mike_mueller.pdf
 * example: http://www.python-academy.com/download/pycon2013/optimization_mmueller.zip

핸드아웃을 그대로 따라해 볼 만함

== Bayesian Statistics Made Simple ==
https://us.pycon.org/2013/schedule/presentation/21/

https://sites.google.com/site/simplebayes/home/pycon-2013

BayesianInference에 대한 새로운 접근. 

== Python for data analysis ==
https://us.pycon.org/2013/schedule/presentation/26/

https://github.com/ContinuumIo/tutorials

numpy, scipy, pandas, disco를 이용한 데이터 분석

=== Disco/Hadoop ===
What is large data? 
 * Can't fit into Excel, R, Memory
 * Can't fit on a single disk - Distributed file stem: SAN, HDF5,…

MapReduce
 * Framework to help solve the problem of distributed computation for distributed data
 * A mass of data: records
 * Split/Map records into key-value pairs
 * Collect/Partition kv pairs (Optional Sort)
 * Buckets are passed to Reduce

MapReduce workflow
 * push code to data
 * lots of network traffic

MapReduce it's a party
 * Disco: Python + Erlang
 * Hadoop: Java-Dumbo
 

== Wiring Up Django Packages ==
https://us.pycon.org/2013/schedule/presentation/11/

https://www.dropbox.com/s/s9b5xmwnl08c9tq/Wiring%20Up%20Django%20Packages.pdf

장고 어플리케이션을 패키지로 배포하는 방법과 규칙들 


= Talks =
== Composability through Multiple inheritance ==
Inter relationship with components

Composability, Compositionality, Quality

lck.django

@llanga, lukasz@langa.pl

== End of object inheritance & Begining of A New Modulity ==
https://code.google.com/a/google.com/p/end-of-object-inheritance
 * Using composition instead of inheritance
 * Composition fails at fault tolerance
 * unidirectional relationship <- bidirectional 

== Things to make test easy ==
유용한 테스트 방법들
 * tempfile.NamedTemporaryFile 을 쓰고, path를 인수로.
 * Other libraries: testfixtures, mock, unittest2

{{{#!Python
from test fixtures import compare, generator
generator(1,2,3)
StringComparison

with self.assertRaises(SomeExcption) as ex:
    #blahblah

from logging import getLogger
from test fixtures import LogCapture
}}}

== Building an image processing pipeline in Python ==
http://endorse.com

pre-processing: cropping(the carpet problem) ext rating lines


== Teaching with the iPython Noteook ==
@jiffyclub http://nbviewer.ipython.org/5165431

pip install ipythonblocks ipython_nose

== Scientific applications in Python ==
http://durden.github.com/python_science_apps

== Customizing the Django Admin The How and the Why ==
Lakshman Prasad @becomingGuru

유용한 방법들
 * Django admin class 에서 form을 정의 
 * BaseModelAdmin 을 서브클래싱~
 * 템플릿 디렉토리 생성 - app_index.html,...
 * Proxy models for aggregation fo data values
 * bootstrap your django admin

django-design-patterns

== Realtime tracking and mapping of geographic objects using python ==
http://gist.github.com/RBURHUM/4559323

 * OpenStreetMap
 * NAIP
 * Map renderers (Mapnik, MapServer, GeoServer)
 * Client-side map js library (OpenLayers) 

A goot real-time tracking solution requires a stateful connection. Polling is BAD.

Celery is awesome

http://abuse.amigocloud.com