[Chennaipy] Minutes of Meeting - March 2015

[Kiran Gangadharan]

Date: 2015-03-29 Sun

* Bitten by Python
  - Speaker: Vijay Kumar B

Vijay recollects how he came across Python while he was still in
college. He had started playing with the language by writing small,
simple
scripts. He talked about how he wrote a Python script for parsing a
log file for debugging an issue on his first day at work. He grieved
about
how indentation came to bite him at the wrong time. He showed a few
code snippets to highlight the mistakes that he had committed. These
mistakes made him rethink as to whether or not Python was "the"
language he wanted to work with.

He highlighted the importance of unit testing and how it helps uncover
all sorts of bugs in code. He also discussed a few unit testing tools
and their possible use cases.

* My Python-BCI Journey
  - Speaker: Kannan

Kannan had attended the meetup for the first time in February. He
mentioned how he was able to easily pickup Python and that being the
reason why he loves the language a lot. He started off by
introducing the concept of BCI and EEG. He showed a clip from NGO's
Brain Games series. He also demonstrated a simple brainwave reading
using
an external kit and the OpenVibe Designer in Windows.

* Python Scripting with SL4A
  - Speaker: Sivasubramanyam

Siva talked about writing simple python scripts on an android device
using the SL4A scripting layer. He talked about how a script he wrote
could potentially replace a timelapse photography app. He gave a short
demo where he was able to listen to an event on a website and trigger
an action on his android device using SL4A.

* How emacs and orgmode replaced a few apps in my workflow
  - Speaker: Kiran Gangadharan

Kiran talked about how he valued privacy and keeping his data to
himself instead of delegating all of it to third party apps like
Evernote and Trello. He demonstrated how he built a workflow to
replace apps like Evernote, Trello, iCal using the famed org-mode in
Emacs.

* Up and Running with Apache Spark
  - Speaker: Krishna Sangeeth

Krishna started out by talking about Big Data and Hadoop and how
Hadoop wasn't the best solution to every problem. He talked about how
inefficient it is to abstract every problem with a mapper and
reducer. He provided a performance benchmark of Hadoop and Spark which
clearly showed how Spark was way more efficient for handling big
data. He also talked about the history of Spark and gave a brief
introduction about it's architecture. At the end, he also provided a
quick demo of using a Spark client in Python for Log Analysis.

* Sentiment Analysis in Simple Steps
  - Speaker: Sharmila G Sivakumar

Sharmila talked about the ease of exploring the NLP capabilities of
Python with a few simple examples of sentiment analysis. She
demonstrated how one could analyse movie reviews and infer whether or
not the movie was worth watching. The examples included how a sentence
could be tokenized, have their terms extracted, filtered to ignore
stopwords(all, just, along, because etc.), analysed to infer the root
word(Lemmatization) and classified into positive and negative reviews
with Naive Bayes Classifier, using the nltk toolkit.

* How to use Slack effectively
  - Speaker: Gaurav Sehrawat

Gaurav gave a quick talk about how Slack based communication is the
future and it's potential benefits over communication via mailing
list. He reminded the audience about the ChennaiPy group on slack and
provided the necessary instructions for obtaining a slack invite.

* Closing Notes

Vijay thanked the speakers, Prof Amritanshu Prasad and the
sponsors. He also talked a bit about the ongoing logo

We had a small discussion about having our meetups on
Sundays, to which there were some mixed reactions. Perhaps, this
could be experimented with, in the future.

Finally, Vijay also suggested an idea of combining a meetup with the
ruby group, and having general talks rather than language-specific
ones. Nothing was finalised in the end.


Cheers,
-- 
Kiran Gangadharan
http://kirang.in