This was my first pycon, I was eager to meet some of the greatest minds in the opensource world. The keynote started with Kushal Das…and he shared some of his wonderful experiences with python. How he began his journey in his undergraduate years and how some of his projects helped the local community educate using computers. He has also been a core contributer to a wide variety of upstream projects.
Upstream a powerful word I came across.
It’s was confusing to decide which talk to attend because they had parallel talks running on 2 auditoriums. So, I had to choose.. They all seem very interesting and most of them were 45min talks.
Without further a do, I decided to go to:
Python + Spark: Lightning Fast Cluster Computing (Jyotiska NK)
Django Design Patterns (Arun Ravindran)
Messing with government data using Python (Anand Chitipothu)
Ya and the lunch was great too.
Meet ElasticSearch - My Pet Crunch Monster (Dhananjay Sathe)
GPU accelerated high performance computing primer (Gurajapu Raja Suman)
The most fascinating talk for me was Messing with government data using Python - Anand Chitipothu
I have some interest in datamining. Mostly I used beautifulsoup and requests… to get things done. Now, I am going to give a shot to scrapy and mechanize. They seem to differ as they can trigger things at runtime.
The talk for elastic search was good too. I am a database newbie. Good to attend some new database stuffs.
My schedule for next day :
The keynote for next day started with voidspace.
> He talked about go lang and his startup juju. And how go error handling differs from others. I also got a chance to talk to him and asked about opensource projects. He adviced me to search for junior jobs on irc channels and see if I can do them.
My schedule was exciting for this day:
Async IO for Dummies (Vajrasky Kok)
Building highly scalable web-services with Gevent - experiences at Plivo (Aditya Manthramurthy)
Faster data processing in Python (Anand S)
Graph everything with Neo4J and Python (Sonal Raj)
I loved this day…I loved Faster data processing in Python - Anand S
The talk was very simplistic. Being a newcomer I understood almost everything. The optimization was a benchmark.
He not only showed ways for faster algorithms, but python libraries to use for faster data processing and also urged people to think whether they need the speed or not.
He used a basic python code to solve the problem, later he used cpython then pypy @jit to optimize the code. It stared with a runtime of 1.5 millisec in standard python to microsec in cpython and to nanosec in @jit.
He optimized his code as fluent as one can get in algorithms.
And yes how can I miss to mention this 9yr prodigy who programmes an arduino.
I will always remember my first pycon.